CN1549171A - Apparatus for realizing high-new technology market fixed standard based on net computation - Google Patents

Apparatus for realizing high-new technology market fixed standard based on net computation Download PDF

Info

Publication number
CN1549171A
CN1549171A CNA031169112A CN03116911A CN1549171A CN 1549171 A CN1549171 A CN 1549171A CN A031169112 A CNA031169112 A CN A031169112A CN 03116911 A CN03116911 A CN 03116911A CN 1549171 A CN1549171 A CN 1549171A
Authority
CN
China
Prior art keywords
market
implement device
new
data
grid computing
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CNA031169112A
Other languages
Chinese (zh)
Inventor
季永萍
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Individual
Original Assignee
Individual
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Individual filed Critical Individual
Priority to CNA031169112A priority Critical patent/CN1549171A/en
Publication of CN1549171A publication Critical patent/CN1549171A/en
Pending legal-status Critical Current

Links

Images

Landscapes

  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The high and new technology market defining standard realizing equipment based on mesh calculation includes: target-oriented market defining virtual and real bourn, cooperative mesh calculation environment, multi-channel data source, attribute characteristic, mode distinction, element marker, scene setting, executing non-linear task, man-machine interacting interpreter, magic engine and visual scene operation. It has also target distinguishing and monitoring module, market defining standard maths model, etc. in the upper part; and artificial intelligent network, real-time information library, etc. in the lower part.

Description

Implement device based on the new and high technology market defining standard of grid computing
Technical field
The present invention relates to the implement device that a kind of new and high technology market is defined, relate in particular to a kind of implement device of the new and high technology market defining standard based on grid computing.
Background technology
New and high technology market defining virtual based on grid computing is the relatively more popular new discipline of current science and technology field with reality, though traditional technical market value assessment increases the pattern that information management system also is based on the computer utility environment with value, but its function and effect are subjected to many-sided restriction, exist many weak points.
Summary of the invention
Therefore, the objective of the invention is to propose a kind of implement device of novelty based on the new and high technology market defining standard of grid computing, utilize the function of grid computing, improved the function and the effect of new and high technology market defining standard and implement device thereof greatly.
The implement device of the new and high technology market defining standard based on grid computing of the present invention comprises: target-oriented market defining virtual and real boundary, the grid computing cooperative working environment, the multi-channel data source, attributive character, pattern-recognition, component identification, sight is set, carry out non-linear task, the man-machine interaction interpreter, evil spirit is carried out the coffin upon burial and is held up, visual field boundary scape operation, above its module, comprise: object identification monitoring and control configuration module, the mathematical model of market defining standard, non-process template, picture of world, intelligence gateway module, below this module, comprising: the artificial intelligence network of grid computing, the evil spirit of visual field boundary scape operation is carried out the coffin upon burial and is held up, the real-time information storehouse, the object resources bank, association's knowledge base, case analysis storehouse etc.
Description of drawings
Describe the present invention in detail below in conjunction with accompanying drawing, in the accompanying drawing:
Fig. 1 is for defining and the synoptic diagram of implement device system based on the new and high technology market of grid computing;
Fig. 2 is the synoptic diagram of the object-oriented visual grid computing and the system integration thereof;
Fig. 3 is the synoptic diagram of visual grid computing pattern;
Fig. 4 is visual grid computing model framework chart;
Fig. 5 defines the synoptic diagram of the non-programmed decision pattern of kernel driving for new and high technology market;
Fig. 6 is the synoptic diagram of the field boundary scape operation of visual grid computing;
Fig. 7 is the synoptic diagram of the deduction of DSP intelligence software chip curing;
The synoptic diagram of the Flexible Control of the non-procedural model that Fig. 8 defines for new and high technology market;
Fig. 9 is the synoptic diagram based on the non-linear task of execution of visual grid computing;
The synoptic diagram of the visual cooperative working environment that Figure 10 defines for new and high technology market;
Figure 11 is the synoptic diagram of the simulated evolution output of visual field boundary scape operation;
Figure 12 is the synoptic diagram based on the integration process of the aggregation operator of visual grid computing.
Embodiment
New and high technology is based upon on the basis of infotech, biotechnology and other discipline developments, is the future of some industry developments, and can change its a lot of industries.New and high technology just at an unprecedented rate create, recombinating industry, discarded traditional method, and produce needs to best new method, core competence and competitive strategy.New and high technology shows right challenge to management through whole tissue, and need comprise many subjects such as finance, management alliance, human resources, the spirit of entrepreneurship, network theory, trade marketing, the market strategy.
In the era of knowledge-driven economy of scientific and technical develop rapidly, market is fast changing, product life cycle constantly shortens, new product is brought in constant renewal in and is regenerated, technical development and the continuous quickening of innovation speed can not all be set up new enterprise at every kind of demand and product, more impossible existing tissue constantly enlarge form large and complete, comprehensive tissue.With minimum flexible cost external demand is made quick tissue and change, realize that the best configuration of resource is the inevitable choice of 21 century enterprise development.
The challenge that new and high technology is brought: define research based on the market of boundary science and disclose the new and high technology possibility until this technology commercialization enters the principal market to the industrialization market management.How to manage the new technology that to create new chance in a large number or destroy existing industry? how to control this complicacy that the field had that is under no restraint, ambiguity and variation rapidly? understand better and how in the industry that is just changed, to compete by new and high technology? traditional thinking, analysis and method for organizing are not enough.Along with the boundary science technology begins to melt the old structure of numerous industries, understand crossing between the fluctuating in market and the various technology barrier, the process itself of making a market strategy just is among the environment of continuous variation and tool height ambiguity.Fast-developing new and high technology has produced huge reaction to the institutional framework in industrialization market, and requires to adopt different management mode and confining method.If new and high technology can realize its potential, it will create very attractive and the market opportunity that allows the people be difficult to ignore.
What hi-tech development caused is chaotic widely distributed.The emergence of internet makes practitioner's anxiety more and more of traditional retail merchant and other industry; The appearance of online household bank makes most bankers feel alarmed when considering to transform banking industry; The great development of biotechnology has been transformed traditional medication chemistry company, forces their reorganization, sets up new alliance, develops new strategy.Some companies more approach the center of the destruction storm that new and high technology causes, but have only Very few companies can avoid the damaging impact of these new energy fully.In one and another industry, the management new technology is just becoming the key of success.
Compared with the functional field of other enterprise, the influences that are subjected to scientific-technical progress of market promotion department more.Therefore, the market management of many industrialization is operated in over the variation and the raising of highly significant in 10 years.And the marketing of new product and the raising of profit on sales become more and more important, the industrialization market process effectively and efficient just more crucial.On each aspect of the industrialization market promotion, all emphasize the use (from the marketing to the highest administration layer) of computer system.The overview of having showed some marketing MIS input, subsystem and output to us.
Marketing MIS input
Marketing MIS (marketing MIS): the information management system of supporting management activitys such as product development, distribution, price policy, sales promotion and sales forecast.Marketing MIS more depends on external data source compared with other technology support system, and these information comprise competition, client, product and have the demand of purchasing power that some internal data source is also very important.
1. market sale target: the marketing relies on the market strategy plan of enterprise about sales target, and the marketing MIS form of such enterprise can be according to the current sale of this market strategy destination organization.Except the market sale plan, also comprise sales force's plan of needs, product/service pricing, channel of distribution, sales promotion and new product characteristic.Enterprise has provided the framework analyzing market information and made a strategic decision in the future market about the market strategy plan of sales target.
2.TPS:TPS comprise a large amount of about market sale and marketing data, comprising the situation of product, client and sales force aspect.Market sale MIS is used for planning and sells or formulate promotion plan with this sample loading mode organizational information.Market sale MIS also can be used for causing the exploitation of new product.Which client is client's sales data can be used to determine to whole profit contribution maximum, also can analyze this data, to develop special client's promotional items, helps marketing personnel to carry out the trial of market and sale.Can also obtain data supervision sales force's achievement by TPS, so that can develop bonus and incentive programme is rewarded the representative of sales ﹠ marketing who has outstanding performance.
3. external resource: competition
Concerning most of enterprises, rival's relevant information all has important value such as quality, packing, sale and the product of new product and service, price strategy, existing product line to distribution of client etc.This information also can be used for the marketing of enterprise's existing product line.The data of relevant competition can obtain from many channels.The employed marketing material of rival, pamphlet, marketing procedure etc. are typical Data Sources.Client and other commercial organization also provide similar data.
4. external resource: market
The another kind of useful oracle of market sale MIS is the market of company's product.Can obtain a large amount of useful data the TPS in the market of having served from company, but but can only from the data source of company outside, obtain data (general picture that functional market survey system is provided can allow company obtain this class data) the further investigation of buyer's behavior and preference in the new markets.
Marketing MIS subsystem and output thereof:
The subsystem of marketing MIS comprises market survey, product development, makes a plan, sales promotion and advertisement, price fixing etc.These subsystems can help marketing management person and administration staff to increase sale, reduce the marketing expenses for their output separately, for the products ﹠ services in future are made a plan, to satisfy the needs that client constantly changes.
MIS is the same with other functionals, and marketing MIS also can support value-added business process.From researching and developing the final distribution of commodity and service, marketing MIS by help to provide better product, quality services, information is so that increase value for client provides the purchase decision guidance to wait timely.Marketing MIS can carry out price analysis, generates corresponding form, to instruct pricing decision.This passes to client by the information aid decision making that marketing MIS provides with direct value.Marketing MIS can also analyze marketing data by help company increases value to satisfy client to the requirement of products ﹠ services.
1. marketing research:
The purpose of marketing research is that the research of standard is carried out in market and customers' preferences.Marketing research can be determined possible client (potential following client), can also find the product that present client really wants or the characteristics of service.Style, color, specification (size), outward appearance, feature such as whether suitable all can be investigated and analysed by marketing research.Can also set price simultaneously, channel of distribution, (quality) guarantee and (after sale) customer service etc.The MIS that markets in case come into the market, the data of collecting from the marketing research project just are used to generate the form of key indexs such as relevant customer satisfaction and overall service level.The form that marketing MIS generates can help the marketing management personnel to go and find out what's going on better, thereby helps tissue to realize its operations objective.
2. product development:
Product development comprises that with feedstock conversion be finished product and service.Emphasis is mainly concentrated on the physical characteristics of product.Many factors are all extremely important in the product development decision-making, as productive capacity, labour, technical merit, engineering factor, material or the like.In many cases, utilize computer program to analyze these factors, select the appropriate combination of labour, material, equipment, engineering design etc.Make or buy decision also can be utilized computer program to assist and finish.
3. place plan:
The problems such as position of place plan process for producing and distribution facility.Where give client, how product is delivered to the key problem that this place is the place analyzing subsystem the product branch.Usually distribution chain finishes to ultimate customers from workshop.The centre is for effectively efficiently products ﹠ services being brought the wholesale and retail network that client adopts.Where be but the best place of settling production facility, setting up distribution center, retail distribution point? the factors such as desired level of production cost, transportation cost, labor cost, localization all are the key factors of this problem.Marketing MIS subsystem can be analyzed these factors, and the best place of determining to lay production facility, wholesale and retail point is to reach minimum cost.The purpose of these place routine analyzers is that total cost is minimized.
4. sales promotion and advertisement:
One of most important function of all marketing work is exactly sales promotion and advertisement.The success that sales promotion of adopting and adline are directly connected to product.When the advertising campaign of which kind of type was taked in decision, the scale of promotion budget, the distribution of budget between each advertising campaign all were very important factors.Owing to utilize computing machine to save time, therefore often utilize the initial budget of computer program formulation and supervise the expenditure of budget and the overall validity of various advertising campaigns.
5. price fixing:
Price fixing is another kind of important and complicated marketing function.Retail price, wholesale price, price rebate all must be formulated.Main factor is exactly the analysis to demand curve in determining pricing strategy, and demand curve has been determined the relation between price and the sales volume.Computer program can help the elasticity of setting price, and formulates various pricing strategies as demand-and-supply-curve of being used for Pricing Analysis etc.Usually, the marketing management personnel can revise those prices that is provided by the computer system of the influence of analyzing price change and gross income being produced.Now, can obtain to feed back fast, thereby make the managerial personnel can be more ever-fasterly than the result who determines pricing decision by computer communication network.This is convenient to formulate pricing strategy more initiatively, and it can be adjusted fast to satisfy market demand.
Sales analysis has contribution to determining which product, sales force and client to profit, and which is not contributed very important meaning.It can also generate some forms and help the marketing management personnel to formulate good Authorize to X.Productized Sales Reports list certain the section during in as one month in, the data of all major products and sales volume thereof.This form will show which product sells well, and which product needed is improved, and which should stop production or stop to be sold.By the Sales Reports of sale personnel classification, list weekly or every month each sales force's total sales volume.
Another importance of marketing function is sales forecast, and it utilizes complicated forecasting techniques and historical data to come forecast demand or marketing activity.These forecasting techniquess comprise trend analysis, moving average, exponential smoothing.Great majority in these technology all need to carry out the lot of data arrangement.Therefore, it is necessary adopting the historical data of complicated computer program analysis usually in formulating sales forecast.
Marketing MIS and subsystem thereof must provide appropriate information in appropriate mode to appropriate people in the appropriate time.The most important internal information source of marketing MIS and subsystem thereof is a transacter.The speed of information and accuracy are very important to the existence of enterprise.Marketing MIS and subsystem thereof provide important resource effectively to formulate these decision-makings to help enterprise by collecting data and generating form.Recognize as early as possible the end what taken place can help enterprise's supervision development trend and carry out function and movable fine setting, thereby finally realize the strategic objective of enterprise.Marketing MIS and subsystem thereof provide different forms to help the supvr for deep understanding of business problem and the market opportunity and correct decision-making.Marketing MIS and subsystem thereof can help the supvr to supervise business procedure so that make better decision-making, seek the value-added new way of service of goods, and help enterprise reaches its target and obtains competitive edge in the market.
Marketing MIS and subsystem thereof are the aggregates that personnel, process, market, performance, database and an equipment are combined, for company manager and decision maker provide the relevant information support, to realize the market performance objective value of enterprise.Most marketing MIS and subsystem thereof are by function of organization's chain building.Typical marketing MIS comprises the management information system of management activitys such as supporting product development, distribution, price policy, sales promotion and sales forecast.Each system is made up of input, processing subsystem and output three parts.
Each enterprise all will define their business objective and service unit, and they also define some assisting workflows, function and are the required action of taking of succeeing usually.The function that some are main: as marketing, Communication with Customer, customer service, merchandise planning, production development, logistics management, financial analysis and cost, risk assessment and channel management etc., all need they to be connected by information resources and analytic process, thus reach to market judge and the service object have one accurately, timely with the supporting technology approach of fully realizing.
Detailed market user's knowledge, and be not only about the transaction and the raw data of financial payments is only that successful enterprise is used for obtaining and client's the key point that keeps getting a profit.Want to create a kind of can share, environment that breakthrough business decision is arranged, just must be converted into the information that to take action to raw data.
On traditional sense; the role of marketing MIS and subsystem tissue thereof just provides TPS or network, technology, information and the flow process of raw data, market proportioning, price policy, merchandise planning and company, so that enterprise goes to collect, manage, protect and provide various data.Yet now, the new function of organizing based on the new and high technology market defining standard and the implement device thereof of grid computing, can not only be converted into the complete information that to take action to raw data, grasped comprehensive market knowledge, and promote as far as possible and be " understanding ", in fact this provided a refinement information and be translated into market and defined and the message structure approach of customers as center.And comprise from all suitable information sources and obtain knowledge, thereby the guiding client uses up-to-date market to concern supporting technology.The most successful market concerns that supporting technology often joins together to set up data warehouse environment with service unit, thereby information is expanded to knowledge, and provides wide road for these very full and accurate Data Views of man-machine interactive access.
The difficulty that defines the new and high technology future market is how to determine the demand of product, and these products are non-existent to the client who is familiar with them not yet.Simultaneously, the speed of the track of technical development and market acceptance also all is unknown.In such market space, the traditional marketing MIS and the disposal route of subsystem thereof have no longer adapted to.But the new function of organizing based on the new and high technology market defining standard and the implement device thereof of grid computing can adopt, and defines the market opportunity under virtual and the actual environment so that can understand and operate in new and high technology market better.In order to make enterprise better adapt to market in the variation, the marketing data warehouse that construction is real-time, collect transaction data from multiple independently channel, then these marketing datas are placed the market subject areas of variation, and then concentrated the readjusting of the information that obtains in this way defined and coordinate.Market subject areas during these change can be any critical thing in enterprise or the industry, as client, product, finance, sale, marketing and distribution point, franchise operation and various channel.The new function of organizing based on the new and high technology market defining standard and the implement device thereof of grid computing, can utilize message structure to be that enterprise creates bigger value: to create a kind of can share, environment that breakthrough business decision is arranged, just must be converted into the information that to take action to raw data.Grasped comprehensive knowledge, and risen to " understanding " as far as possible, just can be in marketing, sale, service, administration, resource management, and accumulation and create wisdom in various levels of decision-making and the planning.
Set up a rational market and define the key of virtual and real support environment, be to allow enterprise diagnosis market and client's behavior, rapidly the market competitive pressure is reacted simultaneously.The supporting technology, operation flow, resource consolidation, Service Management, the channel that define by the market that adapts in changing promote and the data warehouse skill that market is defined virtual data-switching with reality and just can be finished.New and high technology market defining standard and implement device thereof based on grid computing, add that full and accurate is the data source at center with the free market, can make company become quick-witted and elastic on the market, shifty competition player, and can make rational marketing and resource distribution decision-making.
Based on the new and high technology market defining standard of grid computing and implement device thereof (Fig. 1. based on the new and high technology market defining standard and the implement device system thereof of grid computing), can make to be in relation and the interaction that the enterprise of free market in changing started and cultivated higher form between they and service object and the market channel.New and high technology market defining standard and implement device thereof based on grid computing, the competition time domain in free market has been adopted the relation and the interaction of flexibility processing, it can find and analyze the behavior of specific service object and market channel, and makes company therefrom can find initiatively provides personalized product and chance from service to the service object.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, can enlarge by the interaction of supercomputing, dirigibility, creativeness, particularization, customization and complete information, enhancing company in adapt to changing the free market and with the relation of client and market channel.Based on the new and high technology market defining standard and the implement device thereof of grid computing, be to connect and collect in information resources aspect All Activity and the interactive basis of defining in market.Conversion from the trade transactions to knowledge also needs highly intelligentized software to annotate, perhaps Zhuan Ye man-machine interaction modeling technique.As counter or " machine of calculating " (being applicable to the supercomputing center of extensive integrated calculating) are very useful in modeling greatly and fast.For the model that utilizes mass data, the computing power of perhaps relevant with many continuous iterative algorithms normatron is absolutely necessary.Traditional modeling work user imports data and gives model, and model sends to user or model construction person with output again.
(1) often is difficult to obtain or is difficult to produce essential input data or parameter.Such as, lift a common example, inventory model is to be built up on the order equations based of finance, this equation requires to have the cost coefficient that sends a purchase order, but even there is good price to collect and financial system, this coefficient also very is difficult to understand fully.
(2) output that provides of model often is to be difficult to use.The output that the model that moves in computing machine provides generally is the printout of computing machine, even provide with report form, these outputs are also fixed, be difficult to handle.
(3) for complexity, multi-sided problem had proved to be difficult to set up and to keep large-scale comprehensive model already.Long-term development work and lot of data collection work make that spending is very big, and are difficult to original modelling is changed.Common result is that model can only utilize once, or problem is compromised to adapt to existing model.
(4) large-scale complicated model is difficult to make managerial personnel to understand it, thereby they distrust it.When managerial personnel can not make evaluation to imagination, and the time spent of doing that can not understand model, they refuse to rely on large-scale model to do important decision-making.
The way of the continuous maturation that (5) addresses these problems is to utilize the bench model storehouse.Typical situation is that each model is with the data of the form use oneself of oneself, and each model only plays help in a stage of decision process.The policymaker is left in the work of synthetic modeling for or the personnel that deal with problems go to carry out by mental and manpower.
(6) common, minimum interaction is arranged between decision maker and model.But, even under interactional environment, utilize model, interact for data necessary or referential data are provided, to carry out certain and select, the effect of perhaps selecting output format also is very limited.
At first, set up the physical model of system.Need examine the system that analysis is discussed,, omit or simplify some secondary causes, determine the key property and the factor of system, make some hypothesis, be abstracted into the physical model of this system according to the levels of precision of character that studies a question and requirement.
Secondly, on the basis of physical model, set up mathematical model.Need to determine earlier input, output variable and the parameter of model.The input/output variable of model is meant the factor that can get different numerical value that can reflect related characteristics.When determining variable, wish certainly that under the condition that does not reduce degree of accuracy the number of variable is less for well.Usually can handle the quantity that reduces variable like this:
1. between two variablees when similar, but merger is a variable.
2. when variable changed, output (separating) changed little, can be considered as constant to this variable, and getting its mean value usually is this constant.
3. sometimes discrete variable is considered as continuous variable, otherwise sometimes then, continuous variable is considered as discrete variable, may help simplified model like this, be convenient to calculate.
The parameter of model is meant some known facts relevant with the problem of studying.It may change, and also may be indeclinable constant.Object-oriented software can be imagined as the set of some synergistic objects, and these object cooperations provide required service.Analytical model is represented Problem Areas and is designed a model and represents the territory of solving a problem.
The understanding problem
Seek object
Object be grouped into all kinds of in the object-oriented visual programs be based on object technology, and integrate with complete object-oriented development environment.These instruments provide complete unified application and development method: all parts (user interface, Business Entity, computational entity) of developing in each stage all are objects, therefore do not need the conceptual model of problem is mapped to structured representation.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, define an abstract WidgetFactory class, this class declaration an interface that is used for creating the basic window assembly of each class.Each class window assembly all has an abstract class, and concrete subclass has then realized the specific visual sense style of window assembly.For each abstract window assembly class, the WidgetFactory interface all has an operation of returning the new window component object.The client calls these operations obtaining the window assembly example, but client and which concrete class what do not know that they are using be.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, each visual sense standard is all corresponding to a concrete WidgetFactory subclass.Each subclass realizes that those are used to create the operation of the window assembly of suitable visual sense style.The client only creates window assembly by the WidgetFactory interface, and they also do not know which class has realized the window assembly of specific visual sense style.In other words, based on the new and high technology market defining standard and the implement device thereof of grid computing, the client only with the interactive interfacing of abstract class definition, and do not use the interface of specific concrete class.
1. when system will be independent of the establishment, combination of its product and expression.
2. when a system will be disposed by in a plurality of product lines.
3. when you will emphasize the design of a series of relevant product objects so that unite when using.
4. when you provide a product class libraries, and only want to show their interface rather than when realizing.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, create the example of a ConcreteFactroy class usually in the time of running.The product object with specific implementation is created by the concrete factory of implement device.For creating different product objects, the client should use different concrete factories.
The statement implement device is created the operation-interface of abstract good object.
Realize creating the operation of specific product object.
It is interface of series products object statement.
One of implement device definition will be by the product object of corresponding concrete factory establishment.
A concrete factory class only occurs once in an application, and promptly when it was initialized, this made the concrete factory that changes an application become and is easy to.It only needs to change concrete factory can use different product configurations, and this is because a complete product line has been created by an abstract factory, so entire product series can change at once.
The implement device object technology has been contained the whole process from the modeling to the program design.Object technology then is to start with from the centre, from the model of real world, upwards is abstracted into enterprise's level, to the next system programming that refine to.Implement device stresses object model that the method for Problem Areas dynamic perfromance (as incident or user interactions) constructs will comprise different objects with object model by static state or data-driven modeling process (for example by entity is become class entity relationship diagram being become object model) structure.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, really create product and realize by the ConcreteProduct subclass.A most common way is to be factory method of each Product Definition.The concrete factory of implement device will redefine this factory method with appointed product for each product.AbstractFactory is generally each its operation of Product Definition that can produce.The kind of product is coded in the manipulation type structure.Increase a kind of new product requirement and change the interface of AbstractFactory and the class that all are relevant with it.A more flexible but not too safe design is to increase a parameter to the operation of creating object.This parameter has been specified the kind of the object that will be created.It can be a class identifier, an integer, a character string, or other any things that can identify this product.In fact make in this way, AbstractFactory only needs one " Make " operation and indication will create the parameter of the kind of object.
The implement device object model is caught the static nature of problem, and its statement static object structure comprises features such as the attribute of object, relation between objects and class in the system and operation.OMT thinks that object model can obtain by the element from problem statement: noun is possible object, and potential operation or relation represented in verb.Temporary transient, behavior and the feature control aspect of dynamic model expression problem comprise sequence of events, state and the operation of system object.The interaction sequence of OMT between the scene description object.Object restriction or the incident of creating are stored in the occurrence diagram.Dynamic model carries out graph-based by constitutional diagram.A state that object class allowed and a sequence of events in each constitutional diagram display system.
The such static types language of implement device C++ with compare, easier being used in of this variation is similar in the such dynamically typed language of Smalltalk.Only identical abstract base class is all arranged when all objects, when perhaps the pressure that can be requested their client's safety when the product object converted right type to, you can use its parametrization to operate in C++.
The WindowSystem abstract base class defines some interfaces, creates the object of expression window system resource.Concrete subclass realizes these interfaces for certain specific window system.During operation, ET++ creates the example of a concrete WindowSystem subclass, to create concrete system resource object.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, the user can make up a plurality of simple assemblies to form some bigger assemblies, and these assemblies can be combined into bigger assembly again.The Picture class definition polymerization of a Graphic object.The Draw operation of Picture is to call Draw by the subassembly to it to realize that Picture has also realized the operation that some are relevant with its subassembly in this way.Because the Picture interface of implement device is consistent with the Graphic interface, so the Picture object can recursively make up other Picture objects.
Interface of definition is used to visit father's parts, and realizes it under suitable situation in recursive structure.
Expression leaf node object in combination, leaf node does not have child node.
The behavior of definition graphics primitive object in combination.
In the Component interface, realize the operation relevant with subassembly.
Handle the object of combiner by the Component interface.
The user of implement device uses the object in Component class interface and the unitized construction to carry out alternately.
New and high technology market defining standard and implement device thereof based on grid computing, defined the class hierarchy that comprises basic object and compound object, basic object can be combined into more complicated compound object, and this compound object can be combined, and so continuous recurrence is gone down.In the client codes, compound object can be used in any place of using basic object.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, RTL Smalltalk compiler framework has used the Composite pattern in large quantities.RTLExpression is a Component class corresponding to parse tree.These classes are that parse tree has defined a unitized construction.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, Flyweight is a shared object, and it can use in a plurality of field boundary scapes simultaneously, and Flyweight can be as an object independently in each field boundary scape.Flyweight can not make any hypothesis to the field boundary scape that it moves, and the key here is the difference between internal state and the external status.Internal state is stored among the Flyweight, and implement device has comprised the information that is independent of Flyweight field boundary scape, and these information make Flyweight to be shared.And external status depends on Flyweight field boundary scape, and changes according to the field boundary scape, therefore can not share.The implement device user object is responsible for when being necessary external status being passed to Flyweight.
Describe an interface, can accept and act on external status by this interface Flyweight.
Realize the Flyweight interface, and be internal state increase storage space.Must be independent of the field boundary scape of ConcreteFlyweight object.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, create and management Flyweight object.
Guarantee reasonably to share Flyweight.When the user asked a Flyweight, the FlyweightFactory object provided an example of having created or creates a model.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, calculate or store the external status of (a plurality of) Flyweight.
Needed state must be inside or outside when implement device Flyweight carried out.Internal state is stored among the ConcreteFlyweight object; External object is then by Client object storage or calculating.When the user calls the operation of Flyweight object, give it with this state transfer.
The implement device user should directly not carry out instantiation to the ConcreteFlyweight class, and can only obtain the ConcreteFlyweight object from the FlyweightFactory object, and this can guarantee they are suitably shared.
Implement device Flyweight pattern usually and the Composite pattern combine, with hierarchical structure in logic of directed acyclic graph realization of sharing leaf node.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, the behavior pattern of human-machine operation relates to the distribution of responsibility between algorithm and object.Behavior pattern is the pattern of description object or class not only, also describes the communication pattern between them.Have a plurality of objects can handle a request, which object handles should be determined the time of running automatically in request.And the object set that can handle a request should dynamically be specified.
Interface of handling request of implement device definition.
Concrete processor's object on chain is filed a request.
When the client submits a request to, ask along the chain transmission until there being a class object to be responsible for handling it.
The implement device Chain of Responsibility can be simplified interconnecting of object.They only need keep one to point to quoting of its succession, and need not keep quoting of its all candidate recipient.
Be identification request, accessor function of implement device subclass definable is to return such identifier.Perhaps, if the implementation language support, but the type information during recipient's utilization and operation.Member or the soluble order of member view are to carry out an interactive operation, and member and member view can be organized as the structure of hierarchy type.ET++ uses Chain of Responsibility to come the renewal of processing graphics.When a Drawing Object must upgrade its outward appearance a part of, call the InvalidateRect operation.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, OO quasi-mode is to follow the object of decoupling zero call operation and that object with this operation information needed of execution.With the subsystem of high level operation structure that is structured on the primitive operation, quasi-mode provides the method for affairs being carried out modeling.Make that you can be with calling all affairs with a kind of mode.Use this pattern also to be easy to add new affairs simultaneously with expanding system.
The interface of implement device statement executable operations.
Recipient's object is bound to an action.
Implement device is created a specific instructions object and is set its recipient.
The object of quasi-mode call operation with know the object decoupling zero that how to realize this operation.
The specified object of quasi-mode can be handled and expanded as other object.
Implement device can be assembled into a plurality of orders an interdictum duplex.
New and high technology market defining standard and implement device thereof based on grid computing, use the C++ template to realize the object of decoupling zero call operation and have the routine of that object of carrying out this operation information needed, can avoid like this moving and the recipient creates a subclass for each.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, class definition multiple behavior, and these behaviors occur with the form of a plurality of conditional statements in the operation of this class.Relevant conditional branching is moved in their subclasses separately to replace these conditional statements.Need to use the different variants of an algorithm.The algorithm of space/time balance that more self-defined reflections are different.When these variants are embodied as the class hierarchy of an algorithm, can the usage policy pattern.
Implement device defines the common interface of the algorithm of all supports.
Realize certain specific algorithm with the Strategy interface.
Safeguard quoting to the Strategy object.
Interface of implement device definable allows Strategy visit its data.
Implement device disposes with a ConcreteStrategy object.
The Strategy of implement device and the algorithm of Context interaction to realize selecting.When algorithm was called, Context can pass to this Strategy with needed all data of this algorithm.Perhaps, Context can pass to the Strategy operation as a parameter with self.This just allows Strategy can adjust back Context when needed.
Context is transmitted to its client's request its Strategy.The client creates and transmits a ConcreteStrategy object usually and gives this Context.Like this, the client is only mutual with Context.Usually there is a series of ConcreteStrategy class therefrom to select for the client.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, the Strategy class hierarchy is that Context has defined a series of algorithm of Gong reusing or behavior.The succession public function in these algorithms that helps to extract out.As template parameter, in C++, can utilize template mechanism to dispose a class Strategy with a Strategy.Yet this technology only just can be used when the part that cooks noodle satisfies: (1. can select Strategy when compiling.2. it does not need to change when operation.) in this case, the class that be configured is customized for the template class of a Strategy class as a parameter.The use template no longer needs to define the abstract class to the Strategy defining interface.Strategy is also made and a Strategy and its Context can be bound together statically as a template parameter, thereby raise the efficiency.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, a template method defines an algorithm with some abstract operations, and subclass will redefine these operations so that concrete behavior to be provided.The constant part of an algorithm of disposable realization, and leave variable behavior for subclass and realize.Realize the method for a template, the skeleton of a self-defined algorithm.This template method not only calls primitive operation, also call self-defined in subclass operation or the operation in other objects.Template method is a kind of basic fundamental of code reuse.They are particularly important in class libraries, and they have extracted the public behavior in the class libraries.
Realize that primitive operation is to finish step relevant with specific subclass in the algorithm.
Come step constant in the implementation algorithm based on calculating collaborative work.
Subclass can be expanded the behavior of this operation by the operation that redefines parent.
OO operation can be converted to a template method.
The interface that a high level can be provided is to allow its element of this Accessor Access.
Can be a compound or set, as a tabulation or a unordered set.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, use the C++ access control, the primitive operation that template is called can be defined as protecting the member.They are only called this assurance by template method.The primitive operation that must redefine must be defined as pure virtual function.Template method does not need to be redefined.Therefore template method can be defined as a non-empty member function.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, in the interaction modeling, the message sequence by model-dialogue chain is by the user rather than by the model decision, thereby is unfixed for the use of each model.This performance makes the modeling active procedure become independently, and can directly be controlled by the user.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, model bank should have following function:
(1) has the ability that produces new model fast and easily;
(2) has the ability of access and integrated model piece;
(3) have each layer of support user and utilize model that problem is classified, the ability of analysis is divided into uncertain problem the problem of giving a clue;
(4) has model method of attachment, even the ability that model is associated according to knowledge and inference rule;
(5) have calling and carry out the mechanism of model management, carry out the modification of model, the ability of additions and deletions and operation.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, be used for joint between implementation model storehouse and the database:
(1) each model will be obtained whole data values or input data by database.This method guarantees the precision of data, is foundation because the data value of being deposited in the database is validity with data base management system (DBMS) inediting checking procedure;
(2) all model utilizes same database.This characteristics have guaranteed the consistance before and after the data.If a more than model utilizes same data item, so just make that it is identical to all data that application provided.All model utilizes same database can promote the flowability of model.After data were updated, the whole models that are based upon on these data also obtained upgrading simultaneously;
(3) model is sent whole output datas and numerical value back to database.Link between this performance supplies a model, the output that needs another model when a model will be passed through this on-link mode (OLM) during as oneself input.It is also allowed and utilizes dialogue-data chainning to finish display format, and provides output by a model;
(4) model itself adopt one of three kinds of storage formats as database storing in database, this performance makes model more manageability and renewal, because can handle these models by the function among the DBMS.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, model bank is to classify from different perspectives according to the needs of practical application area to set up, and index list by model carries out access.Wherein be listed as routine analytical model (as planning, prediction, statistics etc.) belong to the ripe model of standard, composite construction model (as syntax analysis etc.) belongs to by logic rules and correlation analysis model, and special purpose model (being applied to concrete among the real system and complicated model).
The implement device model bank is stored by the mode of classification hierarchical index, and utilizes the pointer relational links between each level, and points to the restriction of corresponding external environmental factor.The setting of each function of implement device is divided according to the refinement successively and the degree that is mutually related, describe and expression with different angles, at different application corresponding preferential use order and restrictive condition are set, also can directly specify the use order of each model by the user, thereby embody decision maker's part decision-making style, form the expression of priority inferenctial knowledge, after checking effectively, include in the model bank.
The origin of grid computing (Grid Computing) is because of the architecture of this account form as grid.It attempts the computing power of surplus and other idle IT resource relationship are needed the department of high-performance calculation ability within a certain period of time to be supplied to those.The present situation of network development does not also realize sharing of resource aspect at present, has only transmission of Information, so so " embarrassment " can occur: many enterprises every year, the input aspect infotech was all quite huge, but still can not satisfy requirements of one's work.In the unit that has, be equipped with hundreds of platform computing machine, but in the peak period, the usefulness even if 1,000 computers also are nowhere near.And in case design finishes, a large amount of computers can leave unused useless again.
When people obtain information on the internet, be not directly to obtain with internet itself, but must tell computing machine to go to visit some specific websites, this is like in the daily life, we before turning on electric light, tell earlier it we need to obtain electric energy with which power station equally clumsy.The target of grid computing (Grid Computing) allows people use Internet resources simple as electricity consumption exactly.And, increase sharply in case it should be noted that certain tame factory power consumption, can effectively decompose this burden by means of numerous power plant on the electrical network.Effectively solve the experience of load balance on the power network, grid computing just (Grid Computing) is integrated into a virtual supercomputer to the numerous computational resources on the internet, to realize the overall sharing of various resources based on the various resources " the finest fragments of fox fur, sewn together, will make a robe " of CPU.
From this ten years development, be the improvement that CPU, storage, bandwidth all have the very large magnitude, promptly cannot be mentioned in the same breath with software and hardware environment in the past.The development of new technology can enter into a brand-new era in let us promotion internet.In the applied environment of grid computing (Grid Computing), can be utilized the standby time of thousands of station servers, can also rent the people who needs a large amount of computing powers.
New and high technology market defining standard and implement device thereof based on grid computing, distributed neural network is a kind of parallel distributed information Processing Structure, and it is formed some processing units (have local storage and can carry out local information processing capability) interconnection by the one way signal path that is called connection.Each processing unit all has one singly to output to desired connection.The output signal of processing unit can be any desired mathematics type.The information processing of carrying out in each processing unit can be defined arbitrarily under it must be local restriction fully, and promptly its current and processing unit that must only depend on the input signal that processing unit accepts itself is stored the value of remembering.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, a neuron is one n different input, the non-linear unit of n identical output.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, neural network realizes that fuzzy subset's I layer is an input layer, and input quantity x is fuzzy language Ai; The II layer is a hidden layer; III is an output layer, and output quantity is a n dimension fuzzy vector.Each layer I/O function is as follows:
I layer output O (1)=x
II layer output Oj (2)=f (o (1) wj (1)-θ j) (j=0,1 ..., 2n-1)
Wj wherein (1), θ j is adjustable weights, threshold value.F is defined as follows:
f(x)=1/1+exp(-X)
III layer output O (3)J=g (O (2)2jw (2)2j+O (2)2jw (2)2jj=0,1 ..., n-1
Wherein, w2j (1)=1.1, w2j+1 (2)=-1.1, g is defined as follows:
1 x?∈(1,+∞)
g(x)={?x?x∈[0,1]
0 x∈(-∞,0)
With fuzzy language Ai is input, with the fuzzy vector of Ai correspondence (mAi (u1) ... mAi (un)) be desired output, this grid computing can represent fuzzy language Ai (I=1 ..., m).
STEP1: give the connection weights v of input layer unit to the hidden layer unit Ij, the hidden layer unit is to the connection weight w of output layer unit IkAnd the threshold value Q of hidden layer unit jThreshold gamma with output layer unit kCompose the random value of [1,1].
STEP2: for each learning sample pattern to (A k, Y k) (k=1,2 ..., m) carry out following operation:
(1) with A kValue deliver to the input layer unit, again with activation value (output valve) A of input layer unit iBy connection weight matrix V hidden layer unit, produce the activation value of hidden layer unit:
b j=f(∑V ija j+Q j)
J=1 wherein, 2 ... p; F is a s type transfer function.
(2) activation value of calculating output layer:
C k=f(∑W ijb jk)
Wherein, k=1,2 ..., q.
(3) error of calculation function:
ε k=∑(γ k-C k) 2/2
(4) the vague generalization error of calculating output layer unit:
d k=C k(1-C k)(γ k-C k)
Wherein, k=1,2 ... q, γ kIt is the desired output of output layer unit.
(5) calculate the hidden layer unit with respect to each d kError:
e j=b j(1-b j)∑W jkd k
Wherein, j=1,2 ..., p, following formula are equivalent to error back propagation with output layer unit to hidden layer.
(6) adjust the connection weights of hidden layer unit to output layer unit:
ΔW ij=ab jd k
J=1 wherein, 2 ... p, k=1,2 ... q; A is learning rate (0<a<1=.
(7) threshold value of adjustment output layer unit:
Δγ k=ad k
Wherein, k=1,2 ... q.
(8) align the connection weights of input layer unit to the hidden layer unit:
ΔV ij=βa ie j
Wherein, I=1,2 ..., n; J=1,2 ... p; β is learning rate (0<β<1=.
(9) threshold value of adjustment hidden layer unit:
ΔQ j=βe j
Wherein, j=1,2 ... p.
STEP3: calculate the global error function:
E=∑E k
STEP4: if E 〉=ε then repeats STEP2, STEP3, otherwise, change STEP5.
STEP5: study finishes.
With peer-to-peer network, concentrating type supercomputer and to seek the supercomputing center of extraterrestrial life different, new and high technology market defining standard and implement device thereof based on grid computing, the unique distinction of grid computing (Grid Computing) is, can allow a plurality of users to work simultaneously, but not all computing powers are concentrated on the single task role.Realize that information service grid key in application is grid management software.Grid management software can be regarded a kind of middleware as on operating system.In the epoch of Application Gridization, the status of single-set operation system reduces, grid management software is actually higher level grid operations system, its core technology mainly be technology such as incorporate information platform, semantic website, intelligent agent and ontologies (Fig. 2. the object-oriented visual grid computing and the system integration thereof).
New and high technology market defining standard and implement device thereof based on grid computing, in the application of grid computing (Grid Computing), because the demand and supply of information resources is all in dynamic change and be distributed in each different position, server, database or the software of different nodes is called in a service finishing customer requirements possibly, therefore to the response time of server, the bandwidth of network.Particularly the interoperability of various device all has very high requirement on the complicacy of grid management software and dirigibility and the network.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, the genetic algorithm that solves chance constrained programming has been weaved into C programmer.Following example realizes on implement device that all employed parameter is: population scale is 30, crossover probability Pc=0.3, variation probability P m=0.2, the parameter a=0.05 in the evaluation function.
max?x1x2(x3+x4)
{ s.t.
Pr(ε1x 21+ε2x 22+ε3x 23+ε4x 24≤100)≥0.90
X1,x2,x3,x4≥0,
Wherein ε 1 obeys even distribution U (2,3), ε 2 Normal Distribution N (2,1), and ε 3 obeys index distribution EXP (3), and ε 4 obeys the Cauchys C (3,2) that distributes.
Through after 100 iteration, the optimum solution given based on the genetic algorithm of stochastic simulation is: (Fig. 3. visual grid computing pattern)
(x1*,x2*,x3*,x4*)=(3.1751,3.1208,2.2650,1.70280),
Pr{(ε1x* 2 1+ε2x* 2 2+ε3x* 2 3+ε4x* 2 4≤100)≈0.90.
Max f
s.t.
{Pr(ε1x1+ε2x2+ε3x3≥f)≥0.90
Pr(η1x 21+η2x 22+η3x 23+≤8)≥0.80
Pr(r1x 31+r2x 32+r3x 33≤15)≥0.85
X1,x2,x3≥0
Wherein ε 1, and η 1, and r1 obeys even distribution U (1,2) respectively, U (2,3), and U (3,4), and ε 2, and η 2, and r2 is Normal Distribution N (1,1) respectively, N (2,1), and N (3,1), ε 3, and η 3, and r3 is obeys index distribution EXP (1) respectively, EXP (2), EXP (3).
Stochastic simulation will be carried out 2000 circulations each time, and after 2000 generations, the optimum solution given based on the genetic algorithm of stochastic simulation is: (Fig. 4. visual grid computing model framework chart)
(x1*,x2*,x3*)=(105296,0.4226,0.6608),
And desired value f*=2.386 is further, and we have:
Pr(ε1x1 *+ε2x2 *+ε3x3 *≥f *)≈0.90
{Pr(η1x 21 *+η2x 22 *+η3x 23 *≤8)≈0.80
Pr(r1x 31 *+r2x 32 *+r3x 33 *≤15)≈0.85
Lexmin{d1-,d2-,d3-}
s.t.
{Pr{r1x2+r2x2+r3x3+d -1-d +1=10}≥0.90
x1+d - 2-d + 2=4
x2+x3+d - 3-d + 3=2
η1x1+x2+x3≤ε1
Pr{x1+η2x2+x3≤ε2}≥.8
x1+η3x3+x2≤ε3
Wherein parameter ε 1, and ε 2, and ε 3, and η 1 and r1 obey even distribution U (8,10) respectively, and U (10,12), and U (5,7), U (1,2), and U (2,3), η 2 and r2 be obeys index distribution EXP (2) respectively, EXP (3), and η 3r3 is Normal Distribution N (2,1) and N (3,1) respectively.
New and high technology market defining standard and implement device thereof based on grid computing, the applied environment of grid computing (Grid Computing), the various computing resource that is distributed in numerous places will dynamically be integrated, interoperability before the node is inevitable, and this just needs a series of standard agreement and the standard middleware goes working in coordination with, set up, managing and the excavation shared relationship between standard user and the resource.Fields such as the high-end storage of network application, distributed resource load management, collaborative work, data mining had become the battlefield of grid computing (Grid Computing) preliminary examination butcher's knife already.These all demands are used and are made that the industrialization market of grid computing is increasingly mature.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, find that automatically the environment that adapts to is unified modeling language (UML).Automatically find that service can realize automatic monitoring and the identification to workstation, router, system, database, application, the network equipment and their topological structures, and the object storage that will represent these equipment and relation is in the common object repository of grid management software.
The implement device object just can be checked and monitor once definition by three-dimensional real world GUI, visual map browser and other people machine interactive interface.Based on the new and high technology market defining standard and the implement device thereof of grid computing, discovery feature comprises automatically:
The discovery of particular type resource.
The discovery of all resources in the IT facility or the resource discovering of certain subnet.
Operation in the time of a plurality of automatic discovery routine.
Automatically find the automatic startup of service and manually boot.
New and high technology market defining standard and implement device thereof based on grid computing, digitizing real world face comes indicated object user and the employed visual information resource of grid management software by EMPG-4, EMPG-7, EMPG-21 video image processing technology and three-dimension virtual reality technology under unified modeling language (UML) environment, make these resources true more and can manage, simplified the management of distributed resource greatly.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, utilize digitizing real world browser to check that the mode of object is wide-open, and very be easy to customization.Included hundreds of Visualization Model in unified modeling language (UML) environment,, also can under the condition of not revising code, append simply as for some new models as life and work field boundary scape, floor layout or special operation field boundary scape.
New and high technology market defining standard and implement device thereof based on grid computing, Intellective Management Software can be from single reference mark " roaming " whole enterprise, to the present situation of system, network, database, application resource, ontologies, even their physical environment is discerned and is checked.
Visual information resource and the ontologies in the service search network found in the implement device utilization automatically, and they are appended in the common object repository of unified modeling language (UML) environment as managed object.
Utilize the class maker to carry out new class definition, can be described and set up model any entity in the real world.
The formation that implement device can utilize visual information resource browser to check whole real world network is divided the subnet associated therewith and the network segment etc. according to logical relation.
Can be based on the service procedure view of the different customization of demand, resources characteristic, effect, real world field boundary scape and the OO application etc. of different business particular procedure.
Implement device is followed the tracks of the accident of all real world networks in the daily record of incident control desk.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, under unified modeling language (UML) environment, the visual information resource browser of whole real world network can provide the digitizing shape view of object-oriented logical organization.Implement device is by this human-computer interaction interface, the object user can be organized in Drawing Object (such as Background, field boundary scape figure, courses of action figure, outside drawing and hypermedia information view) respectively in the logical layer of real world network, subnet and the network segment according to own personalized service network topology characteristics.
Implement device can show all visual information resources that a certain objects services is used or the specific transactions process is relevant under unified modeling language (UML) environment.For condition and state to object monitor that the implement device user can be provided with trigger and threshold value, be used to tackle the message that this process procedures sends, this can help to solve the early detection and the proactive problem of user profile faulty resource.
New and high technology market defining standard and implement device thereof based on grid computing, the class maker of unified modeling language (UML) environment is a utility easy to use, it can guide the user progressively to finish original class modification in the common object repository, or the increase of new class.The implement device object user can utilize the class maker to come defined attribute, creates the visual touch Lay list that starts application or define the display format of managed object in visual real world network etc.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, under unified modeling language (UML) environment, new class maker can provide view for visual information resource and other storage data that are stored in common object repository.These views can show the visual information of complexity and organize, thereby make the user can be not only simple but also check the service network of own real world all sidedly.All these views can conduct interviews from the information view of visual real world network.
The class browser: the class browser can provide the class of unified modeling language (UML) environment and the demonstration of attribute thereof, comprehensively checks the situation of all classes in the library of object.
Object navigator: object navigator can be listed each unified modeling language (UML) object that environmental classes derived.Utilize this browser can check the information of certain kinds and special object.
The topology browser: topological browser can provide reflection between the managed object dependence and the view of topological relation.It also can show hierarchical relationship and linking relationship between the managed object.
Object view: in common object repository, many managed object representatives all be the information resources view of real world service network.Object pattern identification and other information from intelligent agent also can be provided.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, the incident management of real world service network is the management function to system state/exception.Incident management function by the real world service network, the object user can formulate adaptive strategy, allow unified modeling language (UML) environment that selected message is tackled, automatically it is responded, and needs are caused that the exception condition of paying special attention to sends signal.
Message is responded.
Message is restrained.
Call the order of unified modeling language (UML) environment.
Implement device starts other programs or script.
Implement device is used transmission information to the management of real world service network.
Transmitted message to other by the management service platform.
Implement device is invoked at the order of carrying out on other platforms.
Result to any action makes an explanation, and determines whether action is normally implemented.
Object in unified modeling language (UML) common object repository defines or revises.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, the GUI window that the incident control desk of real world service network calls can monitor the user to institute's generation systems incident.All working procedures and object user's process can directly send the information of query messages to this control desk.Real world view component under unified modeling language (UML) environment relies on its visual information view and digital information source browser, and a highly visual and pattern of real world Service Management intuitively naturally is provided.
Implement device real world view finds to determine the position of services network entity in real world service network facility automatically, as visualization server, digitizing view component, mathematical model of controlled plant, information resource database, and fill common object repository with the object factory of representing these entities and any correlationship.
Unified modeling language (UML) the environment iteration of moving on the network object that the service of implement device intelligent agent real world view gateway is determined to find in the phase one, and with object factory filling (the new class object storehouse under unified modeling language (UML) environment) and the common object thesaurus of representing these agencies and them with the relation of network object.
Implement device common object thesaurus is a warehouse of the setting information of relevant managed entity, their attribute and relation, and the operation of world view is absolutely necessary.The common object thesaurus is filled in the object definition that system can utilize automatic discovery process to collect.DiscoveryMonitor (discovery monitor) provides a visual developing of following the tracks of automatic discovery process.When bringing into operation next subnet or termination search at every turn, automatic discovery just upgrades DiscoveryMonitor.
The thesaurus input of implement device unified modeling language (UML) environment, output are time saving and energy saving practical development sequences, and it allows the object user fill a common object thesaurus with the existing object in another common object thesaurus; With to be stored in this common object thesaurus or another common object thesaurus in the similar object of existing object fill a common object thesaurus.
New and high technology market defining standard and implement device thereof based on grid computing, when the object user starts intelligent agent real world view, open thesaurus input, the output visualization tool of unified modeling language (UML) environment, the alternative user thinks the common object thesaurus of input object, and the object monitor window is revealed.Each managed object all is to have the Managed Object Class of unified modeling language (UML) environment or a subclass of managed object to derive from.The object user can utilize the monitor window of managed object to check and revise the class-based of a managed object and based on the value of the characteristic of example.The monitor window sign of each object is corresponding with a characteristic group of this class.Each field is represented a characteristic, class-based characteristic or based on the characteristic of example.The quantity of the identifier that occurs depends on the class definition of implement device class.
If the object user gives the unique model that Object Creation shows on intelligent agent real world view, must at first use " visual class maker " is class of model definition.When the object user discharged real world view " visual class maker ", implement device class interactive touch device occurred, and showed the hierarchy of available class in the common object thesaurus now.
New and high technology market defining standard and implement device thereof based on grid computing, to on intelligent agent real world view, create and display model, visual virtual class, as seen visual figure is visible and visual virtual, and class-based characteristic must have particular value shown below.Expand class-based characteristic, to show class-based characteristic group, expanding system characteristic then shows the characteristic of this group.
Implement device is after Message Record object and message action object are set up under unified modeling language (UML) environment, they have just constituted the Message Processing strategy, this strategy is specified the incident that needs special processing on the one hand, and must carrying out when incident takes place of task is described on the other hand.When the event message of importing and certain Message Record generation coupling, the processing action that incident management should be taked.The content of this action is from the simple mode display message that just adopts hyperlink visual control process, up to the replying of message, to the processing of problem or carry out some order or other programs etc.The program representation method of model is applicable to the computation model of description schemeization.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, model is non-program mode at intrasystem main storing mode.Based on the model representation of data and can be as the grown form of model based on the model representation of knowledge.In the process of generation model, the technology that adopts quantitative modeling and rational analysis to combine.Quantitative modeling is used for determining the mathematics form of model and related parameter is arranged; Rational analysis is before modeling, in the modeling and all work after the modeling.
The implement device model is created in the man-machine interaction to be finished.It is a dynamic process that model generates.Particularly when the residing environmental change of entity, the build environment of model also should change thereupon.According to the internal cause and the external cause modifications model of system, the process that makes it the tracker motion morphology is called the dynamic generation of model.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, the simulated environment of system is mainly used to simulate the external cause that influences system motion.Quantization factor places database, and non-quantization factor places knowledge base, and the relation between the various factors can be described with model and inference system.
Implement device is when the internal cause of system or external cause change, and structure of models does not change, and the parameter that only changes model is just passable.When the method for the parameter update of model can not improve model for the interpretability of system, should then consider the reconstruct problem of model.
The reconstruct of implement device model is a suitable complicated problems, and different model generating principle correspondences different model reconstructing methods.The content of reconstruct is mainly concerned with dimension, exponent number, constraint condition of model etc., if multiple conllinear problem has taken place, reconstruct will be reselected the substrate of model so.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, the internal cause of system and external cause change not only influences the result of calculation of model, but also influences the rational analysis of model.Particularly when research society, economic phenomenon, the mode of thinking that often forces the decision-maker has taken place to change for policy follows and changes, so the inference mechanism in the implement device model also should be done corresponding adjustment.
The difficulty of new technology future market is how to determine the demand of product, and these products are non-existent to the client who is familiar with them not yet.Simultaneously, the speed of the track of technical development and market acceptance also all is unknown.In such market space, traditional marketing appraisal procedure has no longer adapted to.But, have many methods to adopt in the defining standard of new and high technology market so that can understand market potential under this environment better.
Be verified and have cost benefit in technology, and before market still was in the junior stage, whether market was enough greatly to guarantee that a development plan is still a problem.The track that relevant technologies develops, the validity of standard and support framework is with respect to the target customer's who competes substitution goods income and cost and all necessary consideration of the many imaginations such as common investment of rival in market development.Yet, live through the enterprise of rising in emerging technology market and know that clear and definite answer is unintelligible.This has too many condition and contingency, and these answers are partly to depend on the behavior of enterprise own and try hard to equally answer on the basis of rival's behavior of these problems.The target of a religion reality is to reduce the uncertainty of administration and supervision authorities, obtain simultaneously prior to the adversary take action see clearly.In case reach this target, in emerging chance space, how to obtain and keep a feasible competitive position, these a series of new problems can occur the earliest.
Each emerging technology will be with different prices and VELOCITY DIFFUSION in the market of their expections.Before rising gradually, some market development is very fast, and some market several years that failed.Other markets are before quilt competition technology is shifted the edge onto, and their potential had never been realized.Every kind of approach all is the competitive power results of interaction, and these competitive power can hinder or accelerate the speed of diffusion.Advantage derives from knowledgeable prediction.Promptly view for product is constantly weighed by association from a series of market exploration process, utilizes experience and lessons in each heuristic process to instruct next stage in the development process, shifts to an earlier date the important variation of market prospective simultaneously than the rival.Victors can make chance occur quickly, can invest in the attractive project that has more, and can be with their market, the interests left and right sides.
The ability of the uncertain and expection chance of fast understanding, can by research widely the divergent thinking mode of uncertain things improve, rather than improve by the centralized mode of thinking of seeking a satisfied answer.By with diversified market survey method, different hypothesis, analysis level and Data Source are that the mode at beginning is carried out, and just can obtain these demands better.
The winning relatively intensity that effort promoted that depends on the internal performance of technology and the rival of new technology is provided.Unless the rival is to innovation, the common effort and the investment of marketing and reduction technical costs can elicit latent faculties, otherwise most promising technology all can't find a market.These factors act on the performance footprint that can determine emerging technology together and how soon can meet, and surpass the track of the market demand then.
This process is repeatedly caused by market development one side, and their conviction is that the pioneer is better than market person.Therefore, one is not related to or the future in emergent market will attract the countless losers that reports forever.Each participator might be to the development of technology, and technical equipment is invested, and participates in the project and go, and perhaps these projects can not persuade other participants of same plan fully.By increasing the speed of development of market potential or raising potentiality, the decline of these investment influence and real price is combined, stimulated the development in market.
The process of emerging technology is controlled by rival demand substantially, and it compares the rival and has reached what technology, seeks the new technical advantage that the rival can not imitate easily simultaneously.
The growth of reality and expection in the market demand:
1. the growth of primary market potentiality.
Potentiality economically with statistical expansion.
3. the expansion of market potential (new purposes, new user, better usage).
Perceptible relative advantage:
1. relative worth=discernable income-expected cost.
2. the demand that meets the market segments.
3. perceptible risk.
The excitation of competitive activities:
1. do not use preempting of potentiality.
2. the purpose that obtains and keep on top.
3. the user's resists reaction.
To the most important motivator of market development may be to reduce the real price relevant with substitute.
1. the influence of experience, it is the knowledge by accumulation, economic scale and cause throughput rate to improve and the common result of the technological breakthrough of cost decline.
2. because the influence power of competition makes between general price and the average total cost difference constantly compress.
3. have the method that the user more uses product now by quantity and the encouragement that increases user new on the market, the decline rate of relative price also can have direct influence to the expansion of market potential.
4. new product adopts the client of the leading expection of each stage of process: understanding-understanding-interested-assessment-test-employing.
Commitment in market, individual other enterprise goes to enlarge market and captures market than other rivals with these investments.When development is quickened and competition when strengthening, target diversion obtain or keep an advantage and keep the market share.Yet, all advertising messages, the comprehensive influence that sale is called and commerce is given a performance is being actuated the client and is being passed through the stage of reaction lentamente or promptly.If the expection to emerging technology is fine, investment is huge so; On the contrary, if expection is moderate, or lack confidence, so common investment is appropriate.When the speed that adopts began, Innovation Diffusion depended on the buyer's who experiences the employing process quantity to the speed in market, and buyer's practice of how making decision fast.
For the innovation of noncontinuity, the client of expection is according to its worry and strong degree of demand to risk, by the different phase of oneself selecting to enter.This has caused the difference in period that adopts, and by describing of a period of time, this situation can be represented with a mathematical model.The direct implication of this model is constantly carrying out from a stage form to next stage form, and discontinuous innovation market can be developed.In this, strategy turns to the leader who attempts to become market, and sets up actual market defining standard.In order to carry out the market transition, high-tech enterprise finds that they must determine that on main flow market the specific stage is a target, and develops a supply that increases fully, rather than tries hard to their resource is diffused into many different operational phases.
The process of market understanding is normally by the problem and the chance that occur, development of technology or a kind ofly need have the conviction of more deep opinion to excite to pent-up demand about further innovation.New and high technology market defining standard and implement device thereof based on grid computing, process starts from the information of positive Collection and analysis from the expection client, this information is problem and the demand about the expection client, and definite standard and constraint condition (Fig. 5. the non-programmed decision pattern that kernel drives is defined in new and high technology market).
Based on the new and high technology market defining standard and the implement device thereof of grid computing, understand the market of new technology:
Is 1. (what the notion in emerging technology market to member investigation framework?).
2. collect market information (research and experiment).
3. information dissemination is arrived administration and supervision authorities.
4. explain (reaching a shared intention).
5. utilize market information to make decision.
6. to result's evaluation.
7. the memory of market resource organization and administration.
Before the contradiction confusion of market, can use deviation and incomplete information, but information must be able to be used for explaining so that pattern can be demonstrated and understand.The mode of thinking that influences information is instructing these explanations, those information are through seeking, and selects and has simplified.The market signal of explaining newborn market and emerging market is difficulty especially, because the gerentocratic mode of thinking is incomplete and difference version, often opening an account of expection simultaneously final version from them to being difficult to imagine the original earlier version experience.Can not depend on direct client's feedback, explanation must utilize relevant potential demand, and contextual information is arranged between the trend of long-standing problem and requirement.
The process multilated in many aspects of this market understanding, this illustrates that the ability in each enterprise understanding emerging technology market and the ability of predicting on correct opportunity are very different.The structural obstacle of understanding marketability can be found three long-standing problems.When obtaining information, exist and avoid fuzzy tendency and suppose market intimacy than guaranteeing market for better being inclined to.The experience of accumulation finally is stored in the diffusion memory of market resource tissue, is perhaps just given for change again when demand is arranged.
The utilization of market information is subject to the influence of inertia, and information only just can be used when it is consistent with previous expectation in other words.The market survey Method and kit for only just can be used under the situation of technical abundance being considered to.
When initial stage that methods such as Concept Testing, focus gathering, investigation, experience, analysis-by-synthesis and city's field stimulation are used to indirectly innovation during market, these methods are inappropriate and easily misleading, and this has become a traditional notion.And it is just not strange at all that these methods are used to understand the chance and the strategy that increase innovation on the market of having set up.
This does not also mean that people can not understand only ask for something, application and the attribute of new markets systematically.Existing method need be deacclimatized, and also will change to adapt to distinctive uncertain factor with stylish method.When the object client be no one when knowing, the common sincere advice of being ordered about by the client just has not a particle of value, because this moment, the characteristics that produced by new technology were never experienced in market.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, when the core content that defines when the new technology future market is analysis-by-synthesis and the interactive test findings of experiencing, five concrete analytic models these results that can derive are arranged.When market still is in the junior stage, and the notion of product is still fixedly the time, and it is useful especially analyzing main users and potential demand.Along with the startup in market, diffusion and information aero mode are suitable more equably.
Implement device compares by main users analytic approach and the existent method of analyzing market trend and potential demand, can draw the advantage of main users analytic approach.The plan of most of main users starts from an important trend, and this trend is changing the market that is developed.
Technology is when its developing a certain stage, and main users does not also occur, and perhaps perhaps the most attractive market be different with the market of using this technology the earliest.Sometimes, technology can propose demand that the client does not recognize but that exist.Do you how to recognize this hint and clear and definite these pent-up demands in market?
A method of improving the target market chance is arranged, and Here it is is penetrated into and goes among the client so that find the circumstantial evidence of the market demand.Be to search and understand pent-up demand, can be with " constraint condition that defines of pre-prepd market designs an appropriate methods, and this demand is determined by emerging technology.By potential demand being defined as significantly still still vague implication, remind us to need energy, intuition and acquainted judgement.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, market-oriented problem discovery method can be used for this process of new technology development to break the obstacle of accepting new technology.Another method is that the content that will understand the object client is how they do and how to go truly sensation.Initially coming under the situation of nature that market is defined observed, and can not interrupt movable continuity; Secondly, people can give expression to the non-oral hint of emotion and the spontaneous evaluation that is excited by actual product or prototype; Once more, the observer who was subjected to the special knowledge training under the infotech support environment that defines based on market can find out the indefinite demand that the user fails to perceive and the method for problem of solving.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, in the fuzzy and uncertain market of new technology, if can predict these reflections of future market prior to other people, then the anticipator for the future market must have gained.The effect of market forecast procesures depends on different information sources and to the accurate judgement of important hypothesis.
1. starting stage of the market demand, at this moment product begins diffusion, away from the author of main users and technology.
2. the beginning of cut-throat competition its objective is and catches the most attractive chance.
Because reflection point in market is a competitive power results of interaction of actuating or suppress market development, so the viewpoint of signal that will occur obscuring and contradiction.The prediction of a reflection point is a sixty-four dollar question, and this problem can make feels pattern in omen.This need know note which indication information, and has ability and the wish of isolating transition signal from background interference.This will be with systematic supposition, the combined method of following the tracks of leading indicator and setting up diffusion model.
Technical feature with important parameter that target customer expection is compared on track.
2. in the market exploration stage, main users and other early purchaserses' experience.
3. the client is to the sensation and the risk level of obstacle.
4. in product practicality and market access, the ratio that common investment and competition enter the market.
5. build the process of problems such as infrastructure, solution related standards and supplementary.
The market outlook forecasting institute of neither one new technology can comprise all uncertain contents of relevant object customer response, competitive behavior and technical development, considers that perhaps all complex interactions, discontinuity, initial action and other are non-linear.The thing that the commitment of emerging technology development should be done is that the proof future market has enough greatly to guarantee the development plan of emerging technology.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, the correct starting point of the market outlook foresight activity of new technology should be study from the interaction in market is experienced, and before competition the important reflection point of expection.Define operation in the market that carry out in the field of uncertain most critical, just make and from the interactive experience of the process in market, formed a complete circulation.
1. most of project administrators are found the uncertainty in market is difficult to hold in advance.
2. under the situation that lacks recognizable pattern, use for reference or create the logic that upward reasonably gets on the surface and come decision problem by experience.
3. on the hypothesis basis of trend of market development not being done assurance and not test, make decisions.
4. these hypothesis are hypothesis of the market correct line of the relevant market opportunity and development of new techniques.
5. movable plan of finding based on market and the plot analysis that comes from market are defined in market.
Can enterprise from exploitation and use the emerging technology or the more marketable value of creation from solve known obstacle and prior art.Emerging technology is not the thing that the client will pay in essence.No matter how exciting emerging technology be or have novelty, from client's angle, its value is only from one group of attribute of customer satisfaction.These valuable attribute or advantages, its size are by the technical support of customer response.
The products ﹠ services that enterprise provided have the attribute of converting apparatus.They are some converters, can give the technical capability of enterprise and change into one group of attribute of meeting customer need.Attribute that more hope obtains or undesired attribute still less are provided, can also recognize the demand that will change market to the serious hope of attribute simultaneously.
Expect that market preference to new technology is unusual difficulty, what particularly enterprise faced is the technique variation of high speed development.Yet the Forecasting Methodology of traditional market is inapplicable under such environment.
When a definite feature group became best feature group in a market or in the segmenting market, the market strategy tended to success, therefore will advantageously catch a control ratio of part marketing for enterprise.In definite segmenting market, for accomplishing this, how the restriction that just need understand prior art is to having hindered existing feature group.And the result of market forecast just is present in technology barrier and market is defined in the interaction of environment.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, use new technology to promote an existing interactional technology barrier, the chance that can obtain re-constructing.Define environment by the market of these technology limitation being put into hypothesis, it is clearer that Investment Choice just becomes.Initial, under the technical feasibility condition, the technical development space is very big.Because these combinations are lower than any different price ruling, it can not produce the characteristics combination that attracts clients, then, because after market potential activated by demand, the relatively little change of feature group can produce the big variation on the market.At last, the most attracting combination of market composition may be fully away from existing technology.
Originally many " having destructive " new technology all can not satisfy core market user group's demand.Begin when these emerging technologies come into the market is to be lower than attractive standard that existing rival and market major customer thought.Subsequently, the speed that technical feature is improved makes the feature group fast-developing with emerging technology, might enter and conquer previously can't be approaching customers.
1. these features are distinguishing on market significance.
The feature group how to attract different market user groups (comprise it size, buy the interests of trend, customers).
3. how technology barrier exerts an influence to the interaction between feature and the customers.
In case the supvr that market is defined has understood these feature current values, so next challenge is exactly to understand these features value in future.The supvr that market is defined turns to the demand that satisfies future trend from paying close attention to most popular feature group, estimates the variation on more desired advantage yardsticks in future then conversely.By this process, the supvr that market is defined begins to sketch the contours of the potential direction that strengthens or increase the feature group.
The rival no longer exists, and particularly in highly dynamic environment, entire environment also changes a result who becomes rival's behavior apace.Investment to new technology can be considered to a kind of option on physicals.A small amount of investment during beginning can provide about more investing in the option of technical development or technology commercialization for company.Net present value (NPV) and commutation cash flow method static during following expectancy of determining in the Real Option Theory and traditional financial are analyzed form contrast.In any investment, the company with option on physicals wishes to enlarge the value of oneself investing.On a specific emerging technology, the value of an option is subjected to the influence of the potential size in market that this technology opens.
New and high technology market defining standard and implement device thereof based on grid computing, can attract the analysis of the technology potential that can change a special characteristic and this assessment of importance of particular client group's feature to combine, enterprise just can obtain one more clearly about understanding to this options worth.The supvr that market is defined can be by observing the potentiality of locating the enterprise market or searching potential valuable technology and assess.
Implement device (or expected at least by the risk of framework) again in the market opportunity of seeking again the framework industry, what at first will do is exactly to determine whether new technology has the ability of transmitting prior art feature (or similar feature), and these features will be active in future.Determine that new technology can improve and increase the space of feature, this can cause producing the important change that helps new technology.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, the method that market is defined is to seek the potential feature that emerging technology is created, and determines the market demand of being served by these features then.The net result of an emerging technology is to become certain visual ability of distinguishing the mode manipulation feature.
The cost of breaking a technology barrier is very high, and very big uncertainty is arranged.The application of seeking emerging technology is the process of a confusion, but the major part of this confusion is because the uncertainty of potential market causes.By the uncertainty of understanding potential market and the two aspect contents that solve the market clout of technology barrier, the supvr that market is defined can develop to obtaining some business strategies of maximum effect operation technique resource.Utilize emerging technology to solve obstacle to reach, thereby open the purpose in attractive market opportunity space.
Determine market and relative technology, and evaluate the chance that wherein exists.Produce specific intention, the competitive edge of market potential and company is connected.After the opportunity identification stage finishes, intention is transmitted the physics of interests and the entity that combines is psychologically estimated and improved as one to client, carry out design effort based on this result then.
Deep research will depend on market and how to be defined, the multiple basis of the market segments and specifically indicate the method that market is defined.Sequencing is discharged in these markets, select best the segmenting market to enter.Have high sales potential at those, can have the competition attractive force, can produce scale economics, need investment less and repayment is very big, easier achieving success in the lower market of risk than entering earlier.
Market potential is to weigh according to the sales volume in market and rate of growth.Though the market sales revenue is very important, growth potential also is the key of a new chance of identification.We might find the product chance of a success in a stable market, but sales volume needs directly to capture from the rival there.It is much more difficult that this obtains the market share than the market that is increasing from one usually.
Determine the growth potential in market, need the stage of prediction and understanding life cycle that market is in.Data, services mechanism can provide the data of the relevant total market sales.The expert data service, they can be provided for the data of the rate of growth of prediction markets.After the database of market scale had been arranged, we can use the model of life cycle to predict following growth pattern.Peak and actual data by model prediction match, and on the maximum selling of actual and prediction rational consistance are arranged also.Statistical yardstick is also supported the rationality of this model description.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, this model is based on a kind of hypothesis, and this hypothesis thinks that the first purchase probability of some time is the linear function of former buyer's sum.At first, the probability of purchase depends on the innovator, but after buyer's quantity in the past increased, this probability can increase along with the time.Represent that with the formula form this model is exactly:
P(t)=p(o)+q[Y(t-1)/m]
When wherein P (t)=suppose also takes place to buy, buy the probability of (rate)
The probability of p (o)=initial (buying rate) of selecting
Y (t-1)=, bought the total number of persons of product to the time t-1 end of term
The potential buyer's of m=product total number of persons
The parameter of the spreading rate that q=is estimated
This equation models of implement device is the interactive process of a whole society, and when more people adopted, it just applied the pressure that bigger requirement is adopted.
This model prediction of implement device this model of generation sales volume of product life cycle do not illustrate that product is with old later replacement situation.This model helps to establish the initial growth rate of certain series products, therefore helps improving the new product competitive power when listing just.
This model of implement device is to be used to define the market stage to determine the rate of growth in market rather than the rate of growth of some specific productss.Prediction can be according to these representative values, and perhaps the national kind data of directly making according to early stage time series or statistic processes are estimated.
Implement device utilization model of growth comparison past sales volume carries out simple straight line and extends be eager to excel many.We also should guarantee the determinative of solution curve before predicting, and guarantee that the condition of model hypothesis can be satisfied.
The forecasting research of most life cycles is all for durable consumer goods and industrial goods, but this model also is extended to the influence that technological innovation caused that is applied to look for owing to continuous generation recently.If market is brand-new, and this industry was not also used this model, we will have to according to judgement or carry out the potentiality that market survey comes prediction markets so.In the opportunity identification stage, when only commercioganic potentiality can guarantee required condition always, could use judgement, and then study.
Do like this if the market leader understands market and has the ability, so best preliminary examination location just provides the product that can satisfy most of customer needs.Suppose that the product on the market can be divided into according to customer need " effectively " and " use easily ".Design energy balance validity and the easy product that uses both sides relation, the market orientation of most of customer demands will be selected to satisfy by the company that comes into the market.Being core customer's service (having some companies so does) in market, is to pursue " effectively " or " use easily " and the service of the segmenting market.
The accumulative total sales volume is very important, because it has represented the potentiality of improving product in a lot of industries.These potentiality can with " empirical curve " describe.The unit cost of producing and distributing a product for many manufacturings is along with company's accumulative total is sold being doubled all in decline constantly of grain at every turn.This study phenomenon has been represented the chance that obtains the low-cost competition advantage.
Seldom have for the systematized statistical research of empirical curve to occur, most curves are all depicted from simple cost figure and judgement.Empirical curve exists really, but client is to comfortable, and strong impetus and luxurious preference have surpassed price, become the deciding factor of the market share.
The advantage of cumulative production can come from the market share and the first-to-market of ascendancy.If there is empirical curve, occupying the high market share in a bigger market is the good method that obtains low unit cost.High share has been given relative advantage of company and the ascendancy in the market, therefore also just can control the market strategically.
The potentiality of the scale economics and the market share all are the attractive market characteristics.Because it means that we should concentrate on resource on a few market so that push up sales and reduce cost.Prepreerence should be having the share of capturing market on the market that is increasing of very high potentiality, is possible and enter this market in early days.
Though market increases, the incomer is also few, has scale economics, has antagonistic competitive environment if be in, and product can't obtain the market position of high share, and this market neither a good chance so.
Company can not think that this competitive environment is attractive, unless it has different competitive edges, makes it to succeed in such environment.
1. according to company's situation, enumerate the standard of selecting market and the significance level of evaluating each standard.
2. come to each market scoring according to each standard.
3. calculate the total points that full weight is multiply by in each market scoring.
4. select the best market opportunity according to mark for further studying.The market that score is very low just has been removed.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, need some factors of considering in the market-oriented profile analysis:
1. growth potential: market scale, sales growth rate, life cycle.
2. enter in early days: the priority that enters, user's advantage, the time of deciding.
3. scale economics: win the potentiality of the accumulation sales volume of ascendancy, the importance of empirical curve.
4. the attractive force aspect competition: whether be subjected to rival's attack easily, the potentiality of produce market share, the intensity of competition antagonism.
5. investment: needed investment, raw-material utilizable energy power, necessary technological advancement.
6. repayment: the size in limit, the competitiveness of price structure, rate of return on investment.
7. risk: stability, the possibility of competitive revenge, failed probability, patent protection, the ratio that technology changes, the probability of unfavorable rule.
8. be complementary with the company ability: with the coupling of core technology ability, it is supporting and marketing ability is supporting to have required financial resources and existing product distribution system, can utilize existing sales capacity.
The good market opportunity be those both at the market that should preferentially select, can utilize the chance of company's core technology again.Distribute a relative weighting for main title, this weight is distributed between each ingredient of this title more then.This determines the process of weight to evaluation index, and it is of great value making up and set market priority for company.
In the process of determining weight, maximum results come from the discussion between the executive, and this discussion is guided our notice into various markets, and has guaranteed to pay close attention to these market at first systematically.Many factors are not independently each other, certain seems to make not attractive factor in market, if those abilities that can be converted into competitive edge of energy and company are complementary, this factor may make in fact just in time that this market is attractive so.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, the result of the overall evaluation is that a class value gets the further market and the priority that is used for guiding these researchs of the high potentiality of research.And how to come to determine its each segment market and the border.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, The key factor was that we are used for those indexs in clear and definite market during market was defined.In case we have determined the standard that market is defined, when specifically indicating the scope of new product, it is just very crucial to define market definitely so.Determine the border in the market that they will enter wittingly, and propose the hierarchical system that a cover segments market.Go back at present neither one best or the most correct defining segments market and the method on border, though statistics is helpful sometimes, it still needs gerentocratic judgement at last.
New and high technology market defining standard and implement device thereof based on grid computing, market is defined movable result and has been found the main basis of the market segments, recognize it all is which client is in the target market, to client's situation, their basic need, buying habit is described, and we have also had understanding to the product of vying each other in the segmenting market.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, market is defined provides the framework that carries out the market profile analysis.The profile analysis of this market can filter out less relatively several chances from a large amount of potential market chances.After remaining, select in the market then and further analyze, so that identify the border and the target customers in market.
All markets enter in early days according to growth potential, scale economics, and the attractive force aspect competition, investment, repayment and risk and company's ability standard such as be complementary is estimated.Determined that just can make whole evaluation after the weight of each standard comes.This screening process is to delete those relatively poor market, and concentrates on several markets with high attractive force.These several markets are analyzed determining the border in market, and on the profound level the recognition objective client.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, this information and rate of growth, rate of return, rate of return on investment and risk come together to determine the further priority of research.The analysis of this Market Selection has produced one group of market opportunity, and they have very high attractive force for the new product development of company is made great efforts.
The disturbance of new technology future market and the uncertain chaotic research method that is used to assess existing market.Accept interaction between the speed by market development and market, complicated more to the market evaluation meeting of new technology.When acceptance was accelerated, the improvement meeting of price and performance arrived quickly.Any one that lacks in these key elements all will be slowed down the reception process of technology.
Along with the variation of computational resource and calculating energy, the raising of high performance universal chip cost performance makes the compaction coding method based on software platform have the possibility of practicality; Content-based retrieval and coding are combined consideration, the information of describing video content just should be arranged in packed data, can directly carry out the visit of multimedia messages content at packed data thereby make.This new coding method can have been broken the restriction of original compaction coding method based on specialized hardware based on general-purpose chip, can introduce relate to graphical analysis than complicated algorithm.Based on the image analysis processing technology of the new and high technology market defining standard and the implement device thereof of grid computing, be to support multiple multimedia application (mainly laying particular emphasis on visit) to the multimedia messages content, can come existing field boundary configuration demoder according to the application requirements difference.Coded system is open, can add new effective algoritic module at any time.
In order to support visit to dynamic video content, introduced the notion of visualized objects in the image analysis processing technology based on the new and high technology market defining standard of grid computing and implement device thereof, in fact be exactly to use the delamination area that obtains based on object method, therefore, be based on a class main algorithm in the compaction coding method of content based on the visualized objects method.Be provided at the core technology of the aspects such as effective storage, transmission and operation of video data under the multimedia environment based on the image analysis processing technology (video standard) of the new and high technology market defining standard of grid computing and implement device thereof.Provide support based on the video standard of the new and high technology market defining standard of grid computing and implement device thereof with the form of visualization tool and algorithm, for example effective functions such as the extensibility of extensibility, spatial domain and the time domain of compression, object, error correction to various functions.
Depend primarily on content-based video data method for expressing based on the new and high technology market defining standard of grid computing and the video standard of implement device thereof.Field boundary scape can regard by a series of to have separately the object video of characteristic and form as, and its character comprises shape, motion, texture.This object-based expression is the key that realizes object interaction in multimedia application, and in this application, the user can visit the object of arbitrary shape and operate these objects.
Be based on object based on the coding of the video standard of the new and high technology market defining standard of grid computing and implement device thereof, facilitate operation and controlling object like this, and the conventional compression method be based on frame, obviously can't be to Object Operations.Based on can be based on object in the video standard of the new and high technology market defining standard of grid computing and implement device thereof to bit control, even when low bandwidth, also can utilize code rate allocation method, can be for the interested object of object user with some bit rates of overabsorption, and can distribute some bit rates less for the uninterested object of object user, this sampled images subjective quality just can be guaranteed.
Based on the Object Operations in the video standard of the new and high technology market defining standard of grid computing and implement device thereof the object user can directly be spliced different objects at user side, obtain the synthetic image of user oneself.This can't directly realize in classic method.In expansion, have good dirigibility based on the new and high technology market defining standard of grid computing and the video standard of implement device thereof, can carry out the expansion of time domain and region.And, can expand in time domain or spatial domain according to the objective condition of the existing field boundary bandwidth and the bit error rate.The time domain expansion is to increase frame per second in the enhancement layer on basic layer when bandwidth allows, and can reduce frame per second in basic layer when narrow bandwidth, makes full use of bandwidth to reach, and makes the better purpose of picture quality.The spatial domain expansion is meant carries out sample interpolation to the image in the basic layer, increases or reduce spatial resolution.
(VO, notion VideoObject) realizes content-based expression to have introduced object video in the video standard based on the new and high technology market defining standard of grid computing and implement device thereof.The formation of VO depends on concrete application and the actual environment of living in of series: requiring under the situation of ultralow bit rate, VO can be a rectangular frame (is traditional MPEG, the rectangular frame in H.263), thus with original operating such; For the application that content-based expression is had relatively high expectations, VO may be jobbie or a certain aspect in the field boundary scape; VO also may be the two dimension that produces of computing machine, three-dimensional picture etc.In the video standard based on the new and high technology market defining standard of grid computing and implement device thereof, VO mainly is defined as the different objects that splits in the picture, and each VO has three category informations to describe: movable information, shape letter, texture information.
At first video sequence is carried out the camera lens cutting based on the new and high technology market defining standard of grid computing and the video standard of implement device thereof, each frame in the camera lens is carried out object segmentation, obtain each VO.Based on the encoder functionality in the video standard of the new and high technology market defining standard of grid computing and implement device thereof, at first be the formation (VO Formation) of VO, its less important VO that from original video stream, is partitioned into, be different VO and the three category informations distribution code check of each VO by coding control (Coding Control) mechanism then, each VO distinguishes absolute coding afterwards, and the bit stream complex (MUX) with each VO becomes a bit stream at last.Wherein, can add the mutual control of access customer or control in coding control and compound stage by intelligentized algorithm.
The optimum solution that provides based on the genetic algorithm of stochastic simulation is:
Pr{r1x1 *+r2x2 *+r3x3 *≥10}≈0.90
With
η1x1 *+x2 *+x3 *≤ε1
Pr{x1 *+η2x2 *+x3 *≤ε2}≈0.8
x1 *+η3x3 *+x2 *≤ε3
Ground floor: the node of this one deck directly transfers to down one deck with input value:
F=xi (1),a=f
By formula, the connection weight (wi of ground floor (1)) be 1.
The second layer: each node of this one deck plays the effect of set function.To Gauss's bell shaped function
f=-(xi (2)-mij) 2/ 2ij a=e f
In the formula, mij and ij are respectively the center and the width of j the bell shaped function of I input fuzzy variable xi.Therefore, the connection weight wi of the second layer (2)Play the effect of mij.
The 3rd layer: this layer be connected to fuzzy rule as fruit part, so regular node is carried out fuzzy union:
f=min(x (3)1,…x (3)p,a=f
The 3rd layer connection weight wi (3)Be unit value, as " min " computing in can the replacement formula of the product of fruit part associating.
The 4th layer: for hybrid learning algorithm, the node of this layer has 2 kinds of operational patterns: transmission mode from bottom to top and top-down transmission mode.In transmission mode from bottom to top, the 4th layer connection should be carried out fuzzy " with computing ", sues for peace with the rule that is activated of identical to having " then " part:
f = Σ p I = 1 xi ( 4 ) , a = min ( 1 , f )
Therefore, connection weight wi (4)=1.In top-down transmission mode, the node of this layer is that the 5th layer linkage function is identical with the 2nd layer.For on-line learning algorithm, the node of this layer only moves in the transmission mode from bottom to top.
Layer 5: for hybrid learning algorithm, this layer has two kinds of nodes.First kind of node carried out top-down transmission to training data, with training network.To this node
f=yi, a=f
The 2nd kind of node is to decision signal output execution transmission from bottom to top.The effect of non-Defuzzication is played in XM and the 5th layer the connection that interrelates with it.If these mij (5)And ij (5)Be the center and the width of set function, then Xia Mian function is similar to " center of gravity " non-Defuzzication model 6:
f=∑w?ij (5)xi (5)=∑(mij?ij)xi (5),a=f/∑ijxi (5)
The connection weight w ij of layer 5 (5)Be mij ó ij.For on-line learning algorithm, only there is a kind of node to carry out the transmission from bottom to top of decision signal output.
Coding and Sprite technology have been comprised based on the video standard of the new and high technology market defining standard of grid computing and implement device thereof based on model.After carrying out graphical analysis, investigate each VO earlier and whether meet a model; Consider that again can background adopt the Sprite technology, then background is generated the big figure of a width of cloth in this way, for every frame produces an affined transformation and a positional information gets final product; Just remaining VO is pressed above-mentioned process encoding at last.
Set up the standardized language of a cover based on the new and high technology market defining standard of grid computing and the video standard of implement device thereof: Description Definition Language, in order to declarative description symbol and description scheme, guarantee extendability and long life cycle that it is widely adopted.People can retrieval and index and the audiovisual materials that interrelate based on the video standard data of the new and high technology market defining standard of grid computing and implement device thereof, and these materials can be the multimedia descriptions information that static images, figure, 3D model, sound, speech, video and these elements are formed.Making every effort to based on the video standard of the new and high technology market defining standard of grid computing and implement device thereof can be fast and search out the required dissimilar multimedia materials of user effectively.To carry out standardized description to various dissimilar multimedia messagess based on the new and high technology market defining standard of grid computing and the video standard of implement device thereof, and should describe with described content and interrelate, to realize search fast and effectively.
Mainly be devoted to the information coding of audio-visual data based on the video standard of the new and high technology market defining standard of grid computing and implement device thereof and express, just concentrate in the standardization of the general-purpose interface of the description of multimedia materials (information of expression content rather than content itself).The functional effect of video flowing and video analysis being handled based on the video standard of the new and high technology market defining standard of grid computing and implement device thereof is in the dirigibility of the interactivity of visualized data resource and globalization and data management.
The characteristics and the technology of many association areas must be combined based on the new and high technology market defining standard of grid computing and the video standard of implement device thereof, such as: computer vision, video database and signal Processing etc.The video database personnel pay attention to high-rise description, wish that video standard based on the new and high technology market defining standard of grid computing and implement device thereof provides the structure and the interconnection technique of standard; Signal Processing personnel more pay attention to the analysis of video image with to the understanding of content.
In the built-in function of C language, two subfunctions that produce pseudo random number are arranged.One is
#include<stdlib.h>
int?rand(void)
The random number of its generation is the pseudorandom integer between O and the R_MAX, in stdlib.h, and R_MAX=2 15-1 makes a subfunction be:
#include<stdlib.h>
int?random(int?num)
The random number O of its generation and the pseudorandom integer between the num-1, like this, interval a, the random number on the b} can produce by follow procedure:
U=random (num); //num, an integer is as 10000
u=u/(num-1);
Return a+u (b-a)
If stochastic variable x obeys index distribution, its probability density function is:
f(x)={1/βe -x/β, 0≤x<∞,β>0
0, other,
Be designated as ε XP (β). can produce the random number of obeys index distribution by following method.
Generate u by U (0,1);
x=-β(u);
Return x.
The normal structure of content viewable is provided and has associated mechanism based on the new and high technology market defining standard of grid computing and the video standard of implement device thereof, and to the standardization of content viewable statement, this provides application framework for realizing information retrieval based on contents, and makes to establishment, the exchange of multi-medium data, retrieves and reuse and become more effective.
Demand based on the video standard Description Definition Language of the new and high technology market defining standard of grid computing and implement device thereof: can explain space, time, structure and conceptual relation complicated between description scheme or the description scheme element; Can provide abundant model to connect or one or more descriptions of index and the data that are described; Simultaneously, must be, a man-machine readable platform irrelevant again with application.A kind of language that allows to produce new description scheme (Description Schemes) and descriptor (Descriptors) simultaneously also can be expanded and revises existing description scheme.
Language based on the descriptor (Descriptors) of the video standard of the new and high technology market defining standard of grid computing and implement device thereof, be mainly used to be described below information type: rudimentary audiovisual features, as: color, texture, motion, acoustic energy etc.; Senior semantic feature, as: object, incident and abstract concept; Content-based processing procedure; And media store information or the like.Wish to realize extracting automatically the descriptor information of most low-level features, and when man-machine interactively, can finish senior descriptor information (Fig. 6. the field boundary scape operation of visual grid computing).
The visualization procedure design has been hidden complicated programming procedure by tool using, thereby has simplified the structure work of program design work, particularly graphic user interface.Simultaneously, most of visualization procedure design tools also provide and have the framework of supporting the reusable part library.Visualization procedure design the is convenient customization and the structure (deduction that Fig. 7 .DSP intelligence software chip solidifies) of reusable parts.
A n dimension random vector x obeys the multidimensional normal distribution, and its probability density function is:
f(x1,x2,…,xn)=1/(2π) π/2|∑| 1/2?exp[-1/2(x-μ)’∑ -1(x-μ)].
A n dimension normal distribution random number generative process is:
Generate upper triangular matrix C and make ∑=CC "
Generate μ 1 by N (0,1) ..., μ n;
xk = &mu;k + &Sigma; k I = 1 cki&mu;i , ( k = 1,2 , &CenterDot; &CenterDot; &CenterDot; , n ) ;
Return x=(x1, x2 ..., xn).
At space R nA complex region S in produce even random vector (S is appreciated that into a feasible set in the mathematical programming).
At first, determine a simple region Ω who comprises S, as, a n dimension hypergeometry body
Ω={(x1,x2,…xn)∈R n|ai≤xi≤bi,I=1,2,…,n}.
In the hypergeometry body, random vector relatively is easy to generate.In fact, if xi be [ai, bi] (I=1,2 ..., the n) uniform random variable on, (x1, x2 ..., xn) be exactly uniform random variable on the hypergeometry body Ω.Whether the random vector that generates accepts to depend on it whether in the zone.
The given hypergeometry body Ω that comprises S;
Repeat following process:
By U (ai bi) generates xi respectively, I=1,2 ..., n;
Put x=(x1, x2 ..., xn);
Up to (x ∈ S)
Return x.
Providing a multimedia framework to carry out with numerical information for a user and another user based on the video standard of the new and high technology market defining standard of grid computing and implement device thereof is the interaction of target, and the related each side that interacts is the user.Can limit by five key elements alternately between object user and the numeric item:
1. the content of safety transmits and exchange of value.
2. the visitor easily understands employed project.
3. the personalization of content.
4. in value chain, carry out commercial and useful rule.
5. allow the compatible mutually with it entity of operation in MPEG-4, MPEG-7, MPEG-21 video standard.
To make full use of the related fields of numeric item based on the new and high technology market defining standard of grid computing and the video standard of implement device thereof, but also will contain other media resource simultaneously and describe the description standard form.
1. numeric item is open extendible, can expand in other all media resources and the description scheme.
2. under the situation of structure of not losing subitem and characteristics, project team can be synthesized new project.
3. many group items can the shared cell element.
4. individual element may be by a plurality of local with reference to using in the numeric item.
5. can numeric item be come into force the proper operation numeric item in using.
6. should take open extendible mode to the identification and the correction of numeric item and assembly thereof.
7. clearly define the relation between element and the corresponding description.
8. descriptor can be that simple description also can be whole media component.
9. descriptor also can be described by other descriptors.
10. locator must be stated, allows descriptor relevant with a scope or a specified point of media resource, allows descriptor by locator and media resource backward linkage.
Related needs based on the content representation aspect of the video standard of the new and high technology market defining standard of grid computing and implement device thereof is:
1. data type: should be able to represent large-scale data type, comprise nature and synthetic, also comprise both combinations.As rest image, the video of arbitrary shape, specific and by the 3D model, figure, text, the synthetic audio frequency of nature, nature and synthetic voice etc.
2. content type: can represent the content that all consider data types.
3. validity:, use best significance bit to represent for all data types according to different aimed qualities.Multimedia scene for being made up of the different elements of different types of data should adopt the coding method of different efficient and quality at each element.
4. graded properties: the element in the multimedia scene will have packet performance, and has different extendabilities, on the time, on the space and qualitatively.
5. free access: for the element in the multimedia scene, can be in the time that limits to its different expansion, all elements of different resolution carries out free access.For by different types of data, the multimedia scene that different elements are formed, each element in should the free access scene.
6. wrong the recovery: for the element in the multimedia scene, can protect selected element to avoid producing relevant errors such as channel errors, as the mobile network, atm network or storage medium mistake etc.
7. mutual: as can to carry out alternately with element in meticulous yardstick in time and space and the multimedia scene.For by different types of data, the multimedia scene that different elements are formed should carry out with each element alternately.
8. synchronous: as can to carry out synchronously to all elements and related data in the medium scene.
9. multiplexing: can be multiplexing to the coded data and the related data of different elements in the multimedia scene.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, use object video and synthesize and to synthesize new object video by existing object video.
(1) interval project
Make O=(oid, I are respectively given object video and interval set with I ' v), and relations I are arranged ' C I.In addition, but make that A is the set of an inherited attribute.The video-projection of O on I ' also is an object video O '=(oid ', I ', v '), and wherein oid ' is a new object identifier, and value v ' satisfies following relation:
1) attr (v ')=attr (v) U A, and
2) for each the attribute a among the attr (v ') v ' a=va is arranged.
Interval project is very useful when the new object video of definition one.If existing already present object video will be defined as a new object to the part of the corresponding scene of this object, since the data of description of existing object video can automatically be inherited, then the description amount to new object can reduce.
(2) union operation of object video
Two object video: O1=(oid1, I1, v1) and O2=(oid2, I2, v2) union operation is expressed as O1 U O2, and its result is an object video O=(oid, I1 U I2, v), wherein oid represents a new video object identifier, and I1 U I2 represents two interval unions of sets; And v=[a1:v1 ..., an:vn], ai (1<i<n) should be (vi) among the nattr (v2) wherein at attr
3) the overlapping computing of object video
Two object video O1=(oid, I1 is v1) with O2=(oid2, I2, overlapping operation table v2) is shown O1 n O2, and its result is a new video object O=(oid, I1 n I2, v), wherein oid is the identifier of a new object video, and I1 n I2 represents two friendships after interval the coincidence, and v=[a1:v1,, an:vn], wherein (1<i<n=should be in attr (v1) Uattr (v2) for ai.
To provide following function based on the numeric item identification of the video standard of the new and high technology market defining standard of grid computing and implement device thereof with describing: (Fig. 8. the Flexible Control of the non-procedural model that new and high technology market is defined)
1. accurate, reliable and exclusive identification.
2. do not consider nature, the situation of type and size realizes the seamless identification of entity.
3. the firm and effective recognition method of correlated digital item.
4. the ID and the description of any operation and modification numeric item can both guarantee its security and integrality.
5. handle authorized transactions automatically, content location, content retrieval and content acquisition.
The video data model is the key factor of video data warehouse system operation.The functional requirement of video data modeling:
1. the video model should be supported time series data.
2. data model should be supported sequential branch segment identification at interval.
3. data model should be able to show the relation between the sequential interval.
4. data model should be supported the description to the different levels of each video segmentation.
Because the video data model is can be by unique representative of the video data of computer system processor, obtainable all systemic-functions of object user all depend on the video data model.
An interactively video server system has following characteristics and function:
1. high-performance should allow to carry out simultaneously multi-user access.
2. the video flowing of television-quality is provided.
3. can expand, can be reconstructed for different application requirements easily.
4. portable, portable is (such as Solaris, AIX, HP-UX etc.) on different operating system.
5. Kai Fangxing structure.Allow to use third-party hardware and software.
6. support to comprise video audio frequency, the multimedia database service of image and text.
7. system management facility is provided, comprises user account management, price control, read-write control etc.
8. reliability, for example higher data availability.
9. high performance price ratio.
Oracle Media Server video server system principal feature:
1. modular design comprises directory management, client tracking, bill service etc. and the insertion that allows customized module and third party's module.
2. Gao Du portability.System's support runs on different networks, set-top box, and operating system environment.
3. meet existing multiple interactive television standard.
4. support the parallel supercomputer that cross-platform establishment is huge.It can be a plurality of video flowing services simultaneously, stores thousands of hours video, and can handle thousands of affairs p.s..
5. reliability.Support continual work, in case subsystem fault, multimedia software can guarantee the continuous transmission of data stream.
6. allow the user that Oracle Media Server and existing supporting structure are integrated together and can use third-party monitoring tools.
7. support the multi-medium data service system.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, multi-dimensional database is at the extraction of data and analyzes optimally.Any new data that are loaded into lane database are upgraded usually in batch, and usually from a plurality of data sources.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, set up multi-dimensional database:
1. select you to want to be used for analyzing the business process that is modeled theme.
2. determine the granularity of fact table.
3. distinguish the peacekeeping layer of each fact table.
4. distinguish the tolerance of fact table.
5. determine the attribute of each dimension table.
6. allow the user authentication data model.
Each dimension table of implement device comprises a major key (product, time code, consumer, area code), and the relevant row in the fact table are external keys.Fact table includes a major key (compound), and it is made up of these four external keys.In principle, each external key of fact table must have the copy of oneself in the dimension table.And to have any table of the multi-dimensional database of compound keys must be fact table, this means that each table in expression multi-to-multi dimensional database is a fact table.
Implement device is finished after the fact table, just can determine attribute for each dimension table.How selecting attribute in order to illustrate, is example with the time dimension.The data model of selling business processes is defined as the granularity with astronomical unit relevant with time dimension, so each record of time dimension table is represented one day.Remember: each field in the table defines over the specific sky by the record representative.
When your database server moved on Windows NT, the user can be from catalogue %INFORMIXDIR%/demo/dbaccess visit * .sql and * .unl file.
The implement device user also can clump DB-Access or related object manager the file of command file visit .sql suffix.
(1) initialization is established and is generated number of fuzzy rules t=1, the prerequisite characteristic number I=1 of processing;
(2) for t the neuron in middle layer, calculated characteristics A iThe fuzzy set a of the uncertain correspondence of maximum membership degree value Ir, 1≤r≤k i, satisfy &mu; aij ( w ti ) = 1 &le; j &le; ki max { &mu; aij ( w ti ) } , Recording feature A iAnd fuzzy set a Ir
(3) put i=i+1, and judge,, otherwise change step (4) if i≤n then changes step (2);
(4) the class C of corresponding t the intrerneuron of record HiAnd the connection weight between this neuron and output neuron puts t=t+1 as the probability mean value that activates this neuron sample confidence level, and judges, if t≤k then puts i=1, changes step (2), otherwise continues;
(5) finish.
When the user realized multi-dimensional database, an important step was that exploitation and record load strategy.
In the data warehouse environment of reality, in typical case, the user loads and the output mass data from informix database without LOAD or INSERT statement.
Implement device sales-demo multi-dimensional database uses the data of sales-demo lane database, so the implement device user must create these two databases and realizes the sales-demo database.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, when following two conditions can be used bitmap index for the true time user simultaneously:
1. the key value in the index comprises many duplicate.
2. unnecessary row have index in the table, and optimizer can improve the performance of scan table with it.
● the video interactive store video interactive database internal physical data storage of implement device.
● the video interactive database server of implement device is an application memory service of multimedia data device, simultaneously the method for management database and execution database.
● the environment of the video interactive video client interactive application operation of implement device, it need connect effectively with database server.
● the set of the video interactive class libraries video interactive predefine of implement device and the class that provides, in order to support multi-medium data.
● the video interactive database of implement device is used and used to the video interactive of implement device based on the developing instrument of figure in order to create video interactive.
● the video interactive of implement device is handled the collection of functions of video interactive database, and by C, C++ and other programming languages call.
● the video interactive of Java binding implement device and an interface between the Java, it is java class of video interactive class establishment of each implement device, therefore can use database by Java language.
● the instrument that the video interactive biconditional operation system of implement device works together, it makes supports the Application development environ-ment of ActiveX can use the video interactive database of implement device by Visual Basic or other.
● the video interactive plug-in unit network browser card of implement device, utilize its user to use and to use the video interactive database by the video interactive of World Wide Web operation implement device.
● HTML of WebLink submits program (renderer) to, utilizes it can just can visit the object of the video interactive of implement device without the video interactive plug-in unit of implement device on the standard html page of WorldWide Web.
● the video interactive class set of relational data visit implement device, it makes the video interactive of implement device use and can use and manipulation data on relational database system.
In order to reach the highest performance, the video interactive of implement device has used a kind of distributed client/server structure.Distributed frame makes full use of the power of the modern desk-top computer that has utilized multimedia and user interactions, and server is carried out the mission critical of business procedure in perfecting reliable environment simultaneously.
In the video interactive of implement device, server has very strong processing power, and operation thereon comprises the complicated approach of commercial processes.Transaction management and safety management system are controlled at the method for moving on the server.In addition, operation method has reduced demand to the client application transferring large number of data on server.
The server that is called database engine is being played the part of the role of the reservoir of definition and administrative class definition and object.Database server can also store large multimedia object and other data.
The video interactive client computer is the execution environment of following application: its movingization of control and multimedia demonstration, adjust the interaction of user with external action, and handle retrieval, same server communication for manner of execution to multimedia resource in the database.
The video interactive of implement device provides all standard data type that traditional database had, and a superset that comprises multimedia and other complex data types, for example various images, video and sound have been increased, movingization sequence framework, hypertext and page layout design.
By a multimedia that is included in this software, space and commercial global function collection, the video interactive of implement device is supported these complex data types.By a multimedia that is included in this software, space and commercial global function collection, Jasmine supports these complex data types.Multi-medium data is cubical establish help multi-medium data based on the multidimensional analysis of vision content and the excavation of multiple knowledge, comprise and gathering, relatively, classification, association and cluster.
Related between image content and non-image content feature: belong to this type of as rule " if the first half of photo 50% is blue look, it is likely sky ", because its is associated together content and key word sky of image.
Related with the irrelevant image content of spatial relationship: as rule " if a width of cloth picture comprises two blue look squares, so it also comprise probably one red circular " belong to this type of because related what consider all is image content.
The image content relevant related: belong to this type of as rule " if a red rectangle is between two yellow squares; there is a big oval object so probably below ", because it is associated together object in the image and spatial relationship with spatial relationship.
New and high technology market defining standard and implement device thereof based on grid computing will excavate the association that multimedia object is asked, we can regard each image as affairs, therefrom find out the high pattern of the frequency of occurrences between different images.Under a relatively thicker resolution, excavate the high pattern of the frequency of occurrences, then those images by the minimum support threshold value are done excavation under the further thinner resolution.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, the visualized data excavation is implied from big data centralization discovery with data or knowledge visualization technique and useful knowledge.People's vision system is by the control of eyes and human brain, and the latter can regard the processing and the inference engine of a strong and highly-parallel as, and it has a big knowledge base.Visualized data excavates these powerful assemblies is combined effectively, makes it become an attracting effective instrument, is used for attribute to data, pattern.Bunch and isolated point carry out analysis-by-synthesis.
Data visualization: database can be regarded as with the data in the data warehouse and has different granularities or different abstraction level, also can be regarded as to be combined by different attribute and dimension.Data can be described with multiple visual means, such as box diagram, three-dimensional cube.The DATA DISTRIBUTION chart.Curve.Curved surface.Connection layout, or the like.
Data mining results is visual: knowledge and result that the visual finger of data mining results obtains after with data mining show with visual form.These forms comprise hash figure (scatter plot) and box diagram (by obtaining in the descriptive data mining), and decision tree, correlation rule, bunch, isolated point, generalization rule, or the like.
Data mining process is visual: the various mining processes of the visual formal description of this visual usefulness, therefrom the user as can be seen data from which database or data warehouse, extract, how to extract and how to clear up.Integrated, pre-service and excavation.And, the place and the display mode of the data mining Fang Shi result storage of selecting for use as can be seen.
Interactively visualized data excavates: can use visualization tool in interactively visualized data excavates, help the user to make wise data mining decision-making.For example, the DATA DISTRIBUTION of a series of attributes can be represented (depending on that whole space is to use a circular expression also to be to use the set of row to represent) with color sector or row.This expression mode can help the user to determine which sector at first selected as classification, and which place is best sector cut-point.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, the visual scene operation step of the what comes into a driver's imaging of artificial neural network reduction:
(1) sets up the scenery model according to the fundamental figure unit, and the model of being set up is carried out mathematical description (among the OpenGL point, line, polygon, image and bitmap all as the fundamental figure unit).
(2) the scenery model is placed on the position that is fit in the three dimensions, and viewpoint is set to observe interested scene.
(3) color of all objects in the computation model; Color is wherein determined according to application requirements, determines illumination condition, texture mode etc. simultaneously.
(4) mathematical description of scenery model and color information thereof are converted to the pixel of computer screen, this process is exactly a rasterisation.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, the modeling of visible database figure should be based on the streamline performance structure is optimized.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, chpn structure and what comes into a driver's complexity balance
Main real-time data base Pianli is to reject Pianli and collision detection traversal, and the both can benefit from the database of organizing in the space mode, and they all are based on, and the border body cuts out the what comes into a driver's image.If the hierarchy of a database is with object set close in the space in groups, so can the whole subtree of very simple elimination by the border body of a root node.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, the what comes into a driver's imaging of artificial neural network reduction provides a series of operation balance expense, performance and image quality.Abundant development environment adds that real-time visual maker feature is reduced in order to develop the ideal platform of object-oriented application program the what comes into a driver's imaging.And to the understanding of goal systems graphics architecture and the what comes into a driver's quality and the performance that can make the user obtain to expect towards the design of performance.
These classes of implement device play an important role when creating and managing a new generation's application.When creating various multimedia object, can revise and just use them.During the subclass of designing user definition, they can be used as parent for the demand that satisfies more application-specific.
Also comprise multimedia class and SQL class in the system class that provides except the video interactive of implement device.The multimedia class family that the video interactive of implement device comprises makes video interactive have multimedia function.
Aspect the various multimedia types of storage, multimedia class family has very strong dirigibility.In multimedia class family, main multimedia type all is a class independently, as video and image.
The video interactive of implement device is supported all types of multimedia forms; Yet the structure of multimedia class is easy to be expanded to support in the future available format.
The multimedia that is stored in the video interactive database of implement device can be used figure, video or sound software creation.Easy with the video interactive of the implement device multimedia of packing into as dragging file from desktop and putting into the video interactive object of implement device.In the video interactive database, call ActiveX control or the video interactive by implement device can show multimedia file by Visual Basic.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, the SQL class family by system's support can use, and revises the data in relation and the non-relational database.SQL class family comprises the connection database, revises the method for data and restore data database data.
The third party of the video interactive of implement device develops the function that the partner has expanded the video interactive class libraries.For example, the class libraries of an exploitation has added the function of full text basis and visual inquiry in the video interactive of implement device; Another provides the support to real time data, as the value of financial instrument.
Third party's class libraries has also been expanded the multimedia class libraries of video interactive, makes it support extra medium type, and this type can directly be implemented the video interactive of device and use use.For example, the compressive flow technology of sound and video makes that sound and video can be decompressed when playing.The video interactive of implement device can also managing virtual reality and the video interactive that it adds implement device to used, simultaneously, the digital camera picture shot can be implemented the gamma camera of the video interactive of device and catch and store.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, the class browser is the place of all parts of video interactive database that can see your implement device.Can do following work by the class browser:
● the new class of definition and be the new object of class establishment in database.
● check that in the generic attribute reader attribute reaches the property value of checking an object in the object properties reader.
● create data base querying.
● dilatory object, inquiry or class are on scene.
● in the method editing machine class creation method.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, the video interactive that uses application manager can create implement device is used.The video interactive of an implement device is used and is made up of scene, and scene has same meaning with the Web webpage.
Since can use all database informations by the class browser, you just only need to put into from the dilatory element of wishing of class browser the video interactive scene of implement device so.
When the dilatory element of class browser is put into a scene, you can see a placeholder.A placeholder is a method of representing a database element in the scene.A placeholder is not just represented a static data.Placeholder can be represented dynamic data, as the result of a data library inquiry.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, can check a scene with Design Mode, and can move this scene to test its behavior.When finishing the design that all must scene, you can compile them to create final application in the video interactive application manager of implement device.
After element was added in the scene, they will have vigor because of the behavior of being endowed.Incident can trigger the element in the scene, and action is the behavior that element is produced after receiving incident.Typical action comprises demonstration one frame video, changes scene and sends message to other objects.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, provide one intuitively the interface with the behavior assignment to a scene, and simultaneously without any need for description or coding.
Why Here it is gets next multimedia object and give its mutual behavior from the video interactive database of implement device be very easy.By incidents and the action commonly used of predefine great majority and allow to select them from a drop-down list, the video interactive of implement device makes gives the action perspicuousness of behavior.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, functional mode is described the meaning of the defined Object Operations of action of dynamic model.Functional mode is by data flow diagram, and dynamic modeling illustrates and the behavior of incident trace description object with constitutional diagram, scene.Operation can be from object model attribute derive (read-write properties value and related operation), also can be from dynamically:
Set up the relation between the class
The character and the behavior of definition class
The modeling of object interaction
The state variation of research object
Based on the new and high technology market defining standard of grid computing and the video interactive model of implement device thereof:
Figure A0311691100321
Always have
x max f max f - &GreaterEqual; x max f max f - .
1. put f=+ ∞;
2. generate two clear vectorial ε randomly by the β horizontal cut set of fuzzy vector x and ε respectively 0, x 0
3. if f>f (x 0, ε 0), put f=f (x 0, ε 0);
4. repeating step 2,3 is N time altogether;
5. return f.
New and high technology market defining standard and implement device thereof based on grid computing, object model normally uses class abstract rather than object (being interpreted as the example of class herein) structure, and we we can say that the target of object-oriented analysis is to set up a model based on user, expert and enterprise demand thus.Should comprise in the implement device model the how mutual description of incentive object of the necessary knowledge of relation, each class of satisfying between the required class of application behavior, service that each class must provide and external event about other class (Fig. 9. based on the non-linear task of the execution of visual grid computing).
(1) establishes study number of times t=0, DS=S;
(2) if e ∈ DS item is got in DS ≠ 0, DS=DS-{e}, and carry out following process, otherwise change step (3); If e activates middle layer n 1Neuron, revise weight vector V:
V n1(t+1)=V n1(t)+a(t)·a n1(t)·(cf-V n1(t))
V i(t+1)=V i(t),i≠n 1,1≤i≤k
(3) if a (t)<ε then changes step (4); Otherwise put t=t+1, DS=S, and change step (2);
(4) finish.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, the classification learning process that the intimate human brain of intelligent algorithm deepens constantly is from coarse to fine introduced dynamic competition mechanism in the nerve net of structure.Because the middle layer neuron is to count p from basic class to begin dynamic change, redundant neuron when reaching steady state (SS), network is arranged seldom.Therefore, competition neurons is higher to the classification expression efficiency of experimental data, Knowledge Extraction after being convenient to.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, to liking the abstractdesription of objective world entity.Form by information (data) with to the operative combination of data.Class is the description to a plurality of analogical object denominators.Message is the means of communicating by letter between the object, is used to refer to the operation of object.Method is the description that object receives the action sequence that should take after the message.Example is the concrete object of being described by a certain kinds.
Implement device is built an instance model, and the participant of each use-case is determined on our at first draw border of system then.The participant is outside entity and system interaction time institute role.In case determined the participant, we can to develop the external view how a series of use-case be used for simulation system as the mutual method between participant and system be effectively with setting up prototype system and recording user.
Internal system communication or can not simulated reflection alternately, use-case model is the reciprocation between the descriptive system internal object not.
Conflict between the use-case example can not be reflected that use-case model is not described the conflict and the exception of use-case by model.
Concurrency can not reflect by model, use-case be in essence atom, serial, their parallel situations in can not the reflection system.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, object model has been described the object structure of using, comprising:
The object class of forming application system
Relation between the class
The information of each class (characteristic)
The service that each class provides (public method)
Based on the new and high technology market defining standard and the implement device thereof of grid computing, be the functional relationship of object model inter-entity with constraint definition.Here said entity may comprise object, class, role, attribute, chain with related, constrained the span of entity.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, be the object definition of dynamic model descriptive system following these aspects relevant with the incident or the sequence of operation---incident, sequence of events, the state of definition event context and the tissue of incident and state that sign changes.Dynamic model is described the dynamic behaviour of application with incident trace figure; Incident trace figure shows with constitutional diagram (claiming the Harel constitutional diagram again) how the state of each object is changed by incident under given conditions, and incident trace figure has described the flow of event in using and the variation of Obj State.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, the mathematical programming that has fuzzy parameter can be write as following formation
Wherein x is a decision vector, and ε is the fuzzy parameter vector, and (X ε) goes up objective function to f, gj (X ε) is constraint function j=1,2 ... p, but the mathematical meaning of this fuzzy programming is also indeterminate, and this is to cause not definition of symbol max and constraint because ε is a fuzzy vector.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, Smalltalk is the most representational programming language of object-oriented.All classes of Smalltalk form a level, are made up of root class and a many subclass that is called Object.Class Object provides all objects common behavior, comprises the method for class of Method of printing, tested object of object identifier and the method for copy object.All classes all are the subclasses of Object, and except Object, all classes all have superclass.Each subclass all is based upon on the basis of its superclass, realizes its factum by the method and the instance variable that add oneself.
When if the value of a certain variable program that must remain to is moved in the program next time, its implement device should be stored on the constant medium (the normally medium of program outside), so that can be taken out from medium when needed.The organizational form of external agency depends on the mode that data are used, and it can be an ordinary file, also can be level, netted or relevant database.
1) the object model class is mapped to the non-visualization element of VisualAge.
2) definition non-visual class (common interface) and attribute, action and incident.
3) the mapping object model class concerns and define them in VisualAge.These relations comprise that vague generalization, polymerization are with related.
New and high technology market defining standard and implement device thereof based on grid computing, use a model-view-during effector's model, we we can say and will construct a prototype, and we need use design a model repeatedly class and develop interaction technique for view of visual programming tools.Control structure can be called by message.In Smalltalk, the most basic sequential control structure is represented by a string expression.The control structure of non-order realizes by piece.
1) piece [index ← index+1] is composed to variable i ncrementBlock;
2) piece [sum+ (index*index)] is composed to variable sumBlock;
3) compose to variable sum several 0;
4) compose to variable i ndex several 1;
5) send monobasic message value to piece sumBlock;
6) send binary message * 1 to 1;
7) send binary message+1 to 0;
8) compose to sum 1;
9) send monobasic message value to piece incrementBlock k;
10) send binary message+1 to 1;
11) compose to index 2;
12) send monobasic message value to piece sumBlock;
13) send binary message * 2 to 2;
14) send binary message+4 to 1;
15) compose to sum 5.
Implement device conditional branching similar is in the if-then-else of common higher level lanquage statement.Conditional branching is by realizing to Boolean transmission information.Information identifier is ifTrue:ifFalse:; Wherein argument is two pieces, and message object has only true and false.Object true sends message value to first argument, and falsee is sent message value to second argument.
When causal model changes, implement device the participant take action, and when brainstrust proposed their causal knowledge, the self-adaptation implement device changed its fuzzy causal net.The learning of neuron rule has changed cause and effect rule and limit cycle.The participant has learnt new pattern, has strengthened the memory of old model.In the implement device of complexity, the user can select the dynamic structure of virtual world from a spectrum, and the scope of this spectrum comprises non-linear to mixed and disorderly non-linear from gentleness.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, a virtual world is exactly a power system, when a user or a participant are mobile therein, and the virtual world time to time change.In the simplest situation, have only the user in virtual world, to move.General situation is that user and virtual world all change, and the two mutual change.
Variation in virtual world is inevitable.When the implement device participant was mobile in a virtual world, they impelled incident to take place, and insert new causal model, influence and response old model.Conversely, virtual world also acts on the implement device participant, or acts on its physics or social environment.Based on the new and high technology market defining standard and the implement device thereof of grid computing, virtual world changes its behavior, and can change it self Net of Causation.This feedback causality between implement device participant and its virtual world has generated the power system of a complexity, and this system can modeling event, participant, behavior and data, and this moment, these simulated object were launched in time.
Simple threshold implement device promptly converges on stable limit cycle or fixed point.These limit cycles have been represented " hidden patterns " in the implement device causal net.In a virtual world, limit cycle may be directed according to the order of getting up, work, go home, getting up, the behavior that some are complicated, and for example walking then can be swarmed into simple motion cycle.
Each node in the implement device all makes behavior or target open or close, all himself FCM, Fuzzy control system, oriented target animation system, force feedback or other the input and output mappings of may command of each node.FCM can control the time contact or the time circle in the constructing virtual world.These patterns of implement device have been formed the rhythm of virtual world.
For understand scene and with scene in the object reciprocation, implement device must effectively be represented the three-dimensional data of scene.Three-dimensional scenic represents to include two basic problems: scene rebuilding and scene are cut apart.Scene rebuilding (reconstmction) is meant that (dense depth measurement or sparse depth measurement wide ten calculated the continuous function of curved surfaces, and the wire side that uses many triangular plates or facet sheet to constitute usually in the reality comes approximate representation scene depth measured value from sampled point for use interpolation or approximating method; It is that wire side with the expression scene is divided into several portions that scene is cut apart, and each part is represented an object or a specific zone, helps the accurately realization of post-processing algorithm such as estimation of object identification, curved surface like this.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, based on two frame image estimation disturbance iterative algorithms:
1, initialization depth value { z I, k, I=1 ..., n puts iteration count m=0.
2, under the given depth value according to formula (15.6) estimated parameter
3, according to current estimation and depth parameter, by formula (15.6) calculate corresponding point coordinate (x ' (m) I, k+1, y ' (m) I, k+1)
4, calculate evaluated error
Em = i / n &Sigma; n I = 1 e 2 i
E wherein 2I=(x ' (m) I, k+1, y ' (m) I, k+1) 2+ (y ' I, k+1, y ' (m) I, k+1) 2, x ' I, k+1And y ' I, k+1It is known corresponding point true coordinate.
5, if Em less than predetermined error threshold Er, i.e. Em<Er, termination of iterations then, otherwise, put m=m+1
5, compose a disturbed value to depth parameter.
z (m) I,k←z (m-1) I,k-βe 2 i/z+aΔ (m) I
Wherein a and β are constant coefficients, Δ (m) I=Ni (0, e 2 (m) i) be the zero-mean gauss of distribution function, its variance 2 (m) I=e 2 I
6, got back to for the 2nd step
New and high technology market defining standard and implement device thereof based on grid computing the experiment proved that, this improved iterative algorithm has in the initial depth value under the situation of 50% error, also can well converge to correct kinematic parameter value.
The implement device input imagery can be gray scale image, color image, degree of depth image or their combination.Property detector detects the feature of input imagery, and feature is positioned, and helps hypothesis to generate like this.Object features choose type and the model bank data structure that depends on object to be identified.Suppose that generation module uses image feature to distribute a likelihood value to the object in the scene, this step can reduce the search volume of object identification greatly.In general, model bank is a kind of key map, and it helps removing those impossible candidates from all possible collection of objects, and the hypothesis verification module uses object model to come verify hypothesis, and further provides accurate likelihood value.On the basis on evidence, select for use object as recognition result with maximum likelihood value.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, the algorithm of feature extraction has a lot, according to application, should select reliable characteristic detection method and characteristic positioning method.Many features can be calculated in the 3-dimensional digital image, but they are relevant with the three-dimensional feature of object.Because the characteristic of image generative process, the object features in the scene can be calculated at an easy rate.
Contain bulk information in the various multi-medium datas, understand multi-medium data, must extract its semantic information.Present most multimedia application all was confined to the presenting and represent of multi-medium data, and the rich connotation that does not make full use of media content and had.Information retrieval based on contents, it is exactly retrieval to the multimedia messages content, the reaction media content characteristic that it proposes according to the user search requirement, in jumbo multimedia database, search, the unit returns result for retrieval with the similarity degree that searches requirement according to being retrieved.Traditional database retrieval adopts the retrieval mode based on keyword, has also imported corresponding character descriptor when media data deposits database in, and keyword that provides according to the user during inquiry and descriptor are carried out character match.The defective of this mode is that the semantic feature that multi-medium data has often is difficult to be expressed as certain sign format, and choosing of the keyword of reaction media data characteristics also has very big subjectivity.Based on the new and high technology market defining standard and the implement device thereof of grid computing, information retrieval based on contents is different with traditional retrieval method, and it has merged multiple technologies, comprises image processing, pattern-recognition, computer vision, image understanding, data base administration etc.Information retrieval based on contents is not the content that gets a real idea of medium, but content is discerned and mated, and is needing corresponding domain knowledge in concrete the application.
Visual information is present application prospect media format the most widely.Visual information can be divided into still image and video two big classes.
Color: comprise the composition of color, distribute and mutual relationship.
Texture: comprise the structure of texture, direction and array mode.
Profile: comprise the size of profile, shape etc.
Object: comprise the attribute of subobject in the image, quantity and mutual relationship etc.
The content that the field is relevant: for example resemble in the recognition system five relative positions that close etc. at people's head.
Video is continuous still image sequence:
Camera lens: comprise the switching of camera lens, the division of gradual change etc.
Shooting action: to the extraction of various shooting actions, as shake, push away, draw, follow the trail of etc.
Motion object: search a certain motion object in the video sequence.
Scene: identical camera lens searching and combination etc.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, the complexity of scene depends on that whether the condition (illumination, background, gamma camera parameter and observation point) when obtaining image set up the conditional likelihood mouthful former Zhang Suoshu that speed with model; The condition of scene influences the image of same object significantly, and under different scene condition, the performance of different characteristic detecting device is significantly different.Therefore must consider the characteristic of background, other object and illumination, can obtain effectively and reliable detection to determine which kind of feature.
New and high technology market defining standard and implement device thereof based on grid computing, hypothesis verification is exactly how to use the most possible object of selection in the possible collection of objects of object model from given image, and the existence of each possibility object can prove with their model.We must test the existence that each possible hypothesis is verified the existence of an object or ignored this object.If model is a geometric model, then be easy to gamma camera
Position and other scene parameter are verified object.
The input of implement device vision positioning system is visual 3D sequence of distance and brightness image (ID) sequence, and the model of object generally is three-dimensional model (3n-model), two kinds of orientation problems the most challenging so just occurred.
On the given model any and the current location of model are found out corresponding point in the 3D image.Asking the direct method of corresponding point is to ask the point of minimum distance in the three-dimensional rectangular coordinate space.This and curved surface west: the iterative closest point method in the standard is similar, and on the mathematics, the minimum distance with 3D figure image point y of a bit transferring in the 3D model is defined as
y=arg?min y∈D||x-y||
Wherein, D is a 3D image point set, and the theoretical complexity of search closest approach is 0 (IDI) in the following formula.
If on image last one group of some yi and the model one group, put xi (I=1,2 ..., distance n) all reaches recently, and then image aligns (aligriment) with model.From the initial position of model until aligned position is actually a rigid body translation.Rigid body translation still uses a vector to (R is one 3 * 3 rotation matrix for R, t) expression, and t is a 3D translation matrix vector.Each correspondence provides 3 linear restrictions with respect to positional parameter
yi=Rxi+t
Usually, the 3D data point is subjected to noise pollution
y i=y n i
Wherein β is a 3D variable at random, suppose β obey average be 0 normal distribution (0, σ 2), then for n corresponding point, find the solution positional parameter and become minimum mean-square error is asked minimization.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, constraints module provides the user understands label sets from the marker characteristic value to angle instrument.Constraints module can be used for constructing constraint function.Constraints module provides the interactive interface consistent with other module interface styles.The user can be mutual the shape of appointment classification function, a plurality of classification functions can be made up, and utilize these function systems can generate original hypothesis about minute type objects.The input of constraints module is the eigenwert of extracting, and output is the reliability level of corresponding some target.
Carry out target classification in conjunction with field or user knowledge;
Flag datas such as line segment, zone, surface are carried out cluster, feature association and statistics;
In the low layer of image was expressed, decision was to the area-of-interest of image.
The implement device constraints module provides a subroutine library that is used to visit with the operational constraints function simultaneously.The user can also expand this module by the flag attribute algorithm that adds oneself.In order to distinguish interested object, result after cutting apart is used to compare at first with interested target signature, and the user selects the mark subclass that comprises sample interested, the user generates the classification function of these samples then, and these functions can be used for distinguishing these interested objects at other image.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, when the user roams in virtual environment, in viewpoint and the scene object apart from the time near when far away.When the distance of viewpoint and object was far away, the image of object on projection plane was very little, can use simplified model and was unlikely to influence visual effect.Therefore, can adopt object model to select the foundation of adequate resolution in the simplified model to the distance conduct of viewpoint.
Implement device is in the performance of known computer, after the indexs such as triangle number that can draw in one second kind, can according to the time interval between user's incoming event or space interval calculate for keep current interactive speed the triangle number n/x that should draw, to reach the minimum hardwood speed of Real time dynamic display, per second 10 hardwoods.
Implement device is mapped to one dimension or two-dimensional array to the n dimension input space, constitutes an output region that has significant topological sequences.
The weighted vector that each output layer neuron of implement device interrelates is counted as the result of neuron to the response of input vector.X is marked as with input vector:
x=[x1,x2,…,xp] T
Can be written as with the weighted vector wj of output layer neuron j response:
Wj=[wj1,…wjp] T j=1,2,…N
The triumph output layer is neuronic to be determined, is equivalent to the output layer neuron of selecting weighted vector Wj and input vector x to mate the most.We are optional quantity of stimulus Ij=W t jX is the triumph neuron for maximum output layer neuron.
Step 1. initialization
Optional random value during initialization weighted vector Wj (0).Initial value is selected a little bit smaller usually.Initialization study rate η (0) and field function ∧ I (x)(0).Should get bigger during their initialization as far as possible.
Step 2. is for each vector x execution in step 2a, 2b, 2c in the sample
Step 2a. gets on the input layer that sensory stimuli vector x is sent to network.
The similar coupling of step 2b.:
The neuron of selecting weights vector matching x is as the triumph neuron.Utilization Euclidean rule, win neuronic label I:
I(x)=k||Wk-x||<||wj-x|| j=1,2,…,n
Step 2c. training
The training weighted vector makes that the neuron in activity bubble scope moves towards the input vector direction:
Wj(n+1)={Wj(n)+η(n)[x-Wj(n)] j∈∧ I(x)(n)
Wj (n) other
Step 3. is upgraded learning rate η (n).The linearity of learning rate reduces the gratifying result of generation.
Step 4. reduces field function ∧ I (x)(n).
Step 5. is checked termination condition.
The implement device information retrieval based on contents is a progressively process of refinement.System carries out feature extraction to example, or the feature of describing in the query statement is converted into corresponding query argument.System will search the feature and the feature in the feature database of description and carry out the similarity coupling, and return one group and satisfy the back result of choosing that certain similarity requires.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, ((Data Mining) is meant the process of decimation pattern from data acquisition (model) to data mining.Usually adopt the form of database, so-called useful knowledge is relevant with the demand of particular problem, and offers the user via certain description mechanism.
At given data, model and preferential criterion are sought out specific model and parameter.The task of seeking optimal parameter often is attributed to finding the solution of optimization problem, for example seeks the global maximum of a nonlinear function in parameter space.
New and high technology market defining standard and implement device thereof based on grid computing, a specific data acquisition algorithm is model/criterion/searching algorithm to an example, for example certain data mining algorithm adopts disaggregated model, the method for expressing of model is a decision tree, as preferential criterion, adopt greedy method with the data similarity as searching algorithm.The process of mode discovery is a repeated process, and it at first produces some patterns according to training sample data collection, utilizes parameter data set to select optimal mode then.After determining, pattern utilizes the accuracy of test data set evaluation model.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, integrate module can be integrated relevant feature, can replenish the feature of omission like this and can generate higher abstract representation such as boundary chain code etc.Integrate module can make up the part with similar features (as conllinear, adjacency, color, texture) on the one hand, and the integration of two classes or multiclass feature simultaneously can also use the user to generate hierarchical relationship, thereby clearly expresses complicated structure.
The whole parallel environment of implement device only keeps a data fields, is carried out at after the task division according to new viewpoint at every turn, and the data relevant with each subtask will be sent to corresponding main frame respectively, the heavy distributed process of data fields that Here it is.Most parallel spatial domains volume rendering algorithm all carries out data fields according to this strategy and divides.Adopt this strategy can alleviate the storage pressure of whole cluster environment greatly, however since the spatial domain data fields heavily distribute and must in whole three dimensions, carry out.
New and high technology market defining standard and implement device thereof based on grid computing, a given transaction database, every corresponding collection of data items of affairs, the association that exists between the correlation rule data of description item, promptly, can derive other data item and also in same affairs, occur according to the appearance of some data item in the affairs.
If I={i1, i2 ..., the set that im} is made up of m data item, m is the length of data item collection, length is that the data item collection of k is called k dimension data item collection.The data item that the tentation data item is concentrated is pressed lexicographic order and is arranged, then a k dimension data item collection c can be expressed as c[1] c[2] ... c[k], c[1 wherein]<c[2]<... c[k] .D is an affairs collection, the corresponding data item collection of among the D each affairs, and TCD, each affairs has unique sign TID, to data item collection X CI, claim T to comprise X and if only if X CT. correlation rule is implication X Y, wherein an X CI, Y CI and X п Y=φ, X are called the condition precedent of rule, and Y is called the result of rule.
1. if in affairs collection D, s% is arranged Affairs are supported data item collection X, claim X to have size and are the support of s.
2., claim correlation rule if the support of data item collection XUY is s In affairs collection D, have size and be the support of s.
3. if having the affairs of c% to support data item collection Y simultaneously in the affairs of support data item collection X among the affairs collection D, claim regular X Y in D, to have Degree of confidence for c.
4. ask big data item collection, promptly ask the data item collection of support greater than minimum support threshold;
5. produce correlation rule with big data item collection.The support of data item collection X is designated as support (X). for given big data item collection L, X, Y are the nonvoid subset of L, and XUY=L, X п Y=Φ, if support (L)/support (X) 〉=minconfidence, then generation form is the rule of X Y.
New and high technology city boundary based on grid computing
Figure A0311691100381
Standard and implement device thereof are found from data effectively, the whole process of the novel and pattern that can be understood by the people:
1. application demand analysis.Understand the relevant knowledge of KDD application and user's demand.
2. data are selected.Specified data source and the data subset that carries out the knowledge extraction.
3. data pre-service.The data that step 2 produces are extracted and transformed, and this process will be stored in traditional transaction data storehouse or the data extract in other the information handling system is come out, in addition transition, integrated and checking.
4. data reduction and projection.According to the correlation properties of the target specified data of Knowledge Discovery, utilize the projection operation of database to subdue data dimension, the data volume of utilizing other method for transformation to reduce variable numbers and will operate.
5. determine the target of KDD.According to user's demand, determine the knowledge type that KDD will find.Different types of knowledge need adopt different knowledge discovery algorithms.
6. determine the algorithm of KDD.Preference pattern is found algorithm, comprises determining suitable model and parameter.Selected algorithm is consistent with the target of KDD.
7. data mining.The algorithm that utilization is selected is found the user's interest pattern from given data centralization, and pattern represents with certain specific representation or one group of common type, classifying rules for example, and decision tree returns, cluster, dependence and series model.
8. interpretation of scheme.The pattern of finding is made an explanation, in this process, might repeat some step of front and extract repeatedly to obtain more efficiently knowledge.Visual to the pattern that extracts simultaneously, remove redundant and irrelevant part, help the user to understand the pattern of discovery better.
9. knowledge verification.The knowledge of finding is applied to concrete system, guarantees with the knowledge of finding in the past not inconsistent.
Data pre-service before the implement device data mining step comprises the data selection, cleans, and integrated and checking and interpretation of scheme afterwards all are the committed steps of guaranteeing to obtain useful knowledge.
When the new data source is connected to when being extracted data in data warehouse or the information source and changing, the supervision parts report to integrator with variation.Integrator also comprises several big parts, at first utilizes data modeling tool specified data model, then by extracting the extraction transfer process of conversion component realization from the source data to the data warehouse.Data-switching is to set up a committed step of data warehouse, directly has influence on the quality of data, and it comprises data extract, clean integrated, plurality of sub processes such as checking.It is general that the diversity of data source and complicacy make that extraction/Transformation Program is difficult to.How to realize extracting/the conversion process robotization is the important step that improves the data warehouse versatility.
Implement device extracted data and be converted into integratedly from the affairs type data, the data warehouse of subject-oriented is the process of a complexity, it comprises data read, filters, and is integrated, loads and a plurality of steps such as checking.
Implement device is determined the standard of a metadata, makes the various products in data warehouse field adopt the form and the structure of identical metadata, helps sharing of information.
1. user interface.
2. virtual data base.
3. knowledge I/O storehouse (KIO).
New and high technology market defining standard and implement device thereof based on grid computing can provide following information by user interface:
1. domain knowledge suitably uses domain knowledge extremely important in the Knowledge Discovery process, and is for example visual in rule, and pin subtracts the search volume, improves efficient and the accuracy aspect of finding algorithm.
2. the Data View user is by the data acquisition of data source image file DSM appointment knowledge discovery algorithm operation, and it is the subclass of database global schema.
3. syntactic constraint is by the prerequisite of syntactic constraint user some attribute in can the specific data view as rule, and other are as the conclusion of rule.
4. the different user's interest rule of interest deflection also is different.The interest deflection provides a yardstick of estimating the rule value of being found.
5. threshold value MKS uncertain and support represents knowledge with rule format, and uncertain and support is two standards of evaluation rule.
New and high technology market defining standard and implement device thereof based on grid computing, virtual data base is the interface of MKS and data source, for the application developer of data mining, data source and structure thereof are transparent, the data that VDL user sees are set of tuple. that be associated with VDL is a data source image file DSM, preservation is to be operated the class SQL of data to describe, and utilizing the DSM file can generate Data View .VDL provides one group of function to handle tuple and the attribute that defines in the DSM file.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, the algorithm of intelligence structure is as follows;
1. input pattern is offered network, identification layer selects the maximum of network output as the victor
net j = &Sigma; n I = 1 b ij c i
In the formula, N is a neuron number in the comparison layer.
2. carry out the warning test.And if only if
net j / &Sigma; n I = 1 > p
Then neuron j is by the warning test.P is a warning value in the formula.
If 2a. the victor then shields present victor, and gets back to step 1 and go to select another victor not by test.
2b. repeat this circulation.Up to selecting a victor, forward step 4 then to by the warning test.
3. if do not have neuron to pass through the warning test, then produce a new neuron and admit new pattern.
4. adjusting the neuronic feedforward power of winning plants.The feedback power of the input of renewal from the triumph neuron to it is planted.
Control is weighed the training formula of planting from bottom to top and from top to bottom:
b ij=Lc i/(L-1+∑ck)
In the formula, c iBeing the component of comparison layer vector, is the neuronic numbering of triumph identification layer.
It is W=(w that implement device is established the weight matrix of exporting between layer and the middle layer Ij) K * n, the weight vector between middle layer and input layer is V=(v i) K * 1, learning algorithm be divided into two the step carry out, be described below respectively.
(1) the middle layer neuron number of establishing when initial is k=p, W, and the element initial value of V is composed the random number between [0,1], is expressed as W=(w Ij(0)) K * n, V=(v i(0)) K * 1
(2) establish study number of times t=0, DS=S.
(3) if DS ≠, then get e ∈ DS, DS=DS-{e}, otherwise change step (6).
(4) competitive learning process: import sample e and detect the neuronic state of activation in middle layer to network.Ai during activation (t)=1, otherwise ai (t)=0.Active mode is defined as: get n 1, 1≤n 1≤ k satisfies
D ( X , W n 1 ( t ) ) = 1 &le; i &le; k min { D ( X , W i ( t ) ) }
Wherein D ( X , W n 1 ( t ) ) = | | X - W i ( t ) | | = &Sigma; n j = 1 ( x j - w ij ( t ) ) 2
Claim e to activate middle layer n 1Individual neuron, a at this moment N1(t)=1, a N1(t)=0, i ≠ n 1, 1≤n 1≤ k.If exist a plurality of neurons to satisfy formula (1), can appoint and get one.
(5) weights makeover process: to all i, if 1≤i≤k is W iSatisfy
|D(X,Wi(t))-D(X-W n1(t))|<δ(t)
Then revise weight vector
W i(t=1)=W i(t)+a(t)(X-W i(t))
Wherein a (t) is a learning rate, and δ (t) is the activation field, and both are monotonically increasing function.Change step (3).
(6) if a (t)<ε (<<1) then changes step (7); Otherwise put t=t+1, DS=S
And commentaries on classics step (3).
(7) the neuronic dynamic changing process in middle layer is as follows.
(i) all be in steady state (SS) as if all neurons of middle layer, then each neuronic affiliated class h of mark i, change step (8);
(ii) if all neurons that are in unsteady state of middle layer (are made as n 0Individual) all can not be activated by any sample, then cut down such neuron from network, put k=k-n 0, change step (8);
(iii), then the neuron of middle layer unsteady state is done following processing if above-mentioned condition does not satisfy.
(iv)
(a) if the neuron of a unsteady state is not activated like this by any sample, then cut down this neuron, and put k=k-1;
(b) if the neuron of a unsteady state is worth activation by a certain percentage by the sample more than two classes (being made as the m class), then this neuron is split into m, puts k=k+m-1, and composes initial weight at random for newly-increased neuron.
If the neuron of all unsteady states is all handled, change step (2).
(8) finish.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, when finding the solution complicated optimum problem, each chromosome represents that by a floating point vector its length is identical with solution vector.Here use vectorial X=(x 1, x 2..., x n) represent separating of optimization problem, wherein n is a dimension, then corresponding chromosome also is V=(x 1, x 2..., x n).
The key that implement device is handled constraint condition is all equatioies in (a) deletion constraint condition, (b) the appropriate genetic manipulation of design with the chromosome that guarantees all new generations in feasible set.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, mathematical programming model:
x &Element; max R e - x 2
Feasibility be all real number R.But,, can confirm that optimum solution one fixes on the interval by mathematical analysis, should increase constraint
-5≤x≤5.
The implicit constraint of this class generally is not difficult to find, especially to the problem of management of reality.Therefore, increase implicit condition as much as possible, can accelerate the evolutionary process of problem greatly to reduce the search volume.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, in mathematical programming model, if there are some equality constraints, for example,
h k(X)=0, k=1,2 ... q can be by separating the system of equations that these constraints constitute, with other variable replacement q variable wherein, to eliminate q equality constraint.For example, be without loss of generality, its preceding q variable obtains by the group of solving an equation
x k=h k(x Q+1, X Q+2..., x n), k=1,2 ... q then can be with the variable x in the optimal model 1, x 2... x qRepresent by other variable, so former problem changes into the new problem of having only inequality constrain of one (n-q) dimension.
Implement device is in various spaces, as long as a kind of distance metric of definition can be similarity measurement with the nonincreasing function of this distance metric just.The loss function that minimizes formula (12-22) under the define method that is subordinate to different just obtains different fuzzy clustering methods.
1. set clusters number c and parameter b.
2. each cluster centre of initialization m i
3. the computing below repeating, up to the degree of membership value stabilization of each sample:
Calculate membership function with current cluster centre according to formula (12-25):
Press all kinds of cluster centres of formula (12-24) update calculation machine with current membership function.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, when algorithm convergence, just obtained all kinds of cluster centres and each sample for all kinds of degree of membership values, thereby finished the fuzzy clustering division.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, when having defined dynamic property physics and social environment, just target and purpose have been given for virtual world.Can generate " general an expression " that virtual world is required thus.Implement device can make up simple behavior with simulation " intelligence " behavior.In a nested implement device, each node can be controlled the simple FCM of himself successively.Complex behavior as walking, occurs from the network of simple reflex, and the available two-value limit cycle of nested simple realization device is a finite state machine net with this process simulation.
Implement device limit or rule can be mapped to another with a sub-notion, and this a little notion mapping has formed a set that input is mapped to fuzzy system or fuzzy " if first " rule of output.Each mapping all is the fuzzy rule or the state space rule patch that connect fuzzy set.In the input-output state space, patch has covered the figure of certain function.Fuzzy system is average with overlapping patch, to generate approaching of a continuous function.Based on the new and high technology market defining standard and the implement device thereof of grid computing, how sub-notion can map to different reactions.This just produced one to more various reaction of changing in the virtual world (Figure 10. visual cooperative working environment that new and high technology market is defined).
The implement device mobile participant in virtual world who strengthens, the two-value output state of this implement device has moved the participant, and each FCM state is mapped as mobile equation or the function that approaches.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, for the two-value limit cycle among the relational matrix E is encoded.The method of TAM is with the addition of the relevant battle array of the weighting in the continuous state, for limit cycle C1 → C2 → C3 → C1 is encoded, at first each 0 usefulness-1 will be replaced, thereby each two state of value Ci is converted into a secondary state vector Xi, this moment, E was weighted sum:
E=q 1X T 1,X 2+q 2x T 2,X 3+…+q n-1X T n-1X n+q nX T n,X j
The length of limit cycle should will take place otherwise disturb less than the quantity of notion.The right suitable weight of each relevant battle array can be improved coding, and can strengthen the memory space of FCM.
If a FCM can enough more nonlinear mathematics methods change its node and limit, it just can simulate these complicated virtual worlds, and the cost of paying is exactly that virtual world may become the chaos virtual world with unknown balance behavior.The user may want these novel results are added virtual world, so that its more attractive.Perhaps, the user can select a virtual world, and it is that appropriateness is nonlinear, and the cycle equilibrium state is arranged.Another virtual world that extremely then is the user selects has strong nonlinear and has only aperiodic equilibrium state, can be for the example of explanation at random the virtual visual field boundary scape operation that can change by wish ground of gravity potential (Figure 11. the simulated evolution of visual field boundary scape operation is exported).
Based on the new and high technology market defining standard and the implement device thereof of grid computing, by reasonable division and the tissue to calculation task and processing node, this class algorithm can effectively reduce data traffic; Because each grouping computing power is pre-estimated,, still can be obtained satisfied load balance though the task division mode is static.Its shortcoming is that the data redundancy amount is excessive, and in fact the storage of total system consumes the scale that not only depends on raw data, and is directly proportional with the number of group.
Implement device is synthetic to the resampling and the image of volume data.Whole algorithm is carried out according to divide-and-conquer strategy, and the resampling process is divided finally through recurrence and finished by each processor is parallel, and image is synthetic then to be undertaken by trace-back process.The data of this algorithm are divided and visual synthesis strategy is well-designed, so very high parallel efficiency and extensibility are arranged.
(1), can reduce the public part between the image subsection of each merging for the relatively-stationary environment of viewpoint.For this reason, can be along near a coordinate axis of direction of visual lines, data fields is divided into a plurality of sections (slice) orthogonally.
(2) for cartoon scene, whole data fields can be organized according to the k-D tree construction along three coordinate directions.In addition, can also by pre-service respectively in the specified data field the big and gradient of gradient and then adjust the density that data are divided on each direction near zero scope.
(3), also need consider between each processor difference at aspects such as computing power, communication capacities for isomerous environment.
Each processor of implement device obtains corresponding image subsection by after the data fields subclass (such as section) that is assigned with is resampled, and image is synthetic to combine all these according to succession exactly in image, obtain final drawing result.This algorithm has adopted two fens exchange (binary-swapping) technology, visual building-up process is walked abreast finish.Shown in, to from the w width of cloth image subsection of N processor and row image synthetic, can be by the log between the processor 2N
Based on the new and high technology market defining standard and the implement device thereof of grid computing, the partial results by each minute process calculating gained in the concurrent program need be merged to obtain a complete result with aggregation operator.N process P1 wherein ... Pn is undertaken by carrying out interaction code C alternately.Claim that this n process is mutual litigant (or participant).Can carry out after all participants have arrived C if code C must work as, then journey is synchronous alternately.If a process can be waited for other processes after reaching C, and carry out C, then claim this alternately for asynchronous.
Mutual during N=2 is called the mutual of two litiganies, and be mutual if n is called many litiganies greater than 2.When a process when another process sends message, it is point-to-point communication that two litiganies' communication just is called.It is mutual that many litiganies often are called set alternately.
When process P of implement device runs into an interaction code C,, just can obtain multiple interactive mode if the entrance and exit condition is provided with distinct methods.Have in n member mutual at one, if interactive mode can be described as the process index simple function, then claim this to have mode of rule alternately.Mode of rule is meant, count i and j for two given indexs, if there is an algorithm, it can be at a fixed time with the space in determine i process whether influenced the pattern of j process (Figure 12. based on the integration process of the aggregation operator of visual grid computing).
According to the requirement of systemic-function,, provide two-dimensional image to handle the interface and browsed two-dimensional image, and carried out the extraction and the editor of outline line thereon for the user based on the new and high technology market defining standard and the implement device thereof of grid computing.Simultaneously, also provide three-dimensional image to handle the interface and generated three-dimensional image, and carried out various operations for the user.Handle in the interface at two-dimensional image, provide to enter the button that three-dimensional image is handled the interface.Equally, in the three-dimensional interface, also provide the button that returns two-dimentional interface.But, because the function of these two modules is relatively independent, so, when design, also stay interface, make and can make two mutual incoherent autonomous systems from now on easily into.
The major function that 3-D view is handled is as follows:
1. the description document of transferring person/close: the choice box that opens file, select File, transfer person simultaneously first image and outline line file thereof, and in Information display file information.
2. store description document: deposit the current modification that outline line file, volume data file etc. is made in description document, after having stored, can change the corresponding information in the Information.
3. a slice image under showing: a slice image under will in drawing is distinguished, transferring person, the interior information accordingly of the Information of the outline line file of a slice image under the mediator, and change simultaneously.
4. show to go up a slice image: with show that a slice image class seemingly down.
5. a slice image that shows appointment: eject dialog box, the sheet of certain sheet number in the input chip sequence, the respective profile of then transferring person line file is revised the corresponding information in the Information simultaneously.
6. amplification region function: pin key in the mouse (can drag) in the left window master draws the district, will show a red line frame in this draws the district, image that the red line frame is interior and outline line etc. will be shown blown up in the amplification region of right window.In the amplification region, can carry out the operation that all carry out in the district that draws.Pixel in the grid representative master drawing district that bottom shows in the amplification region, a pixel of relevant position in the red line frame during just the main drawing of grid representative is distinguished.What arbitrary place chose in this grid in the mouse point all is main same pixel of drawing in distinguishing.
7. the outline line file of transferring person/close.
8. storage wheel profile file: the broken line of storage wheel profile is represented to represent with B-spline curves simultaneously.Outline line also can be represented by a bankbook line.More than two the step all can eject file selection box, with import file name.
10. at the outline line that shows a slice image on the background: show the outline line of going up a slice image with another color, this outline line can only be seen, can not edit.Purpose is for editor's participate-reform to some extent during outline line on this film image, reduces up and down too big error between two outline lines.Shown and also can close behind the outline line of last a slice image and do not show.
10. add/delete bar outline line one by one: after choosing the button that adds outline line, can eject a dialog box, import the class-mark of initiate outline line.
is point of adding/deletion in current outline line.
edits the point on the outline line.
becomes the B-SPL with current outline line from segmented fitting.
measures distance between two points: mainly be internal diameter in order to survey blood vessel etc.If in first, imported scaling point (being equivalent to have engineer's scale) this moment, then measure the back report physical length that finishes, unit is (units).Otherwise report number of picture elements, unit are (pixels).
measures two wire clamp angles.
The input of  engineer's scale: this is because of the image of scanning input computer to the end of cutting into slices from entity, therebetween through repeatedly conversion, in order still to embody actual size, on first, an engineer's scale is arranged, and generally shows the length of 1cm.Calibrate (being certainty ratio) at these 2 during the user can put, thereby follow-up measurement etc. can be carried out according to actual size.
The input of calibration point.
Figure A0311691100422
Calibrate the outline line in current all cross-sectional images.
Outline line information is changed into volume data and storage.
Figure A0311691100424
Withdraw from.
In addition, the implement device mouse can be finished hit detect function.If during any entity, then do not have feedback signal in.If the point in the point on the current outline line, then can delete, operation such as editor, if the limit on the current outline line in the point, operation such as then can add some points thereon.If non-current outline line in the user point, then with this outline line as current outline line.Then, relevant putting and the various operations on limit can deserved to carry out on the front-wheel profile.
1. the volume data of transferring person file: eject file selection box, mediator's desired body data file.Utilize the MarchingCubes algorithm to generate the dough sheet collection of inhomogeneity (being different structure), and give their priming colors and transparency.Provide the interface with input sample density during generation.
2. the dough sheet that simultaneously shows one or more structures according to the color and the transparency of appointment.
3. to the three-dimensional dough sheet collection that shows be rotated (around transferring axle and Y-axis), scaling or to flexible (being equivalent to change the sheet spacing).
4. around the rotation Z of a certain direction to flexible and can both return to original state individually to translation.
5. opacity, the class name (as bone, muscle etc.) of a certain class dough sheet are set on control panel, and open palette color is set.
6. on palette, there is multiple mode that color is set.
7. the dough sheet collection is carried out smooth operation (smooth), to strengthen effect.
8. whole dough sheet collection of storage generation comprise all kinds of colors, opacity and class name.
9. store the dough sheet collection of a certain class, so that access separately from now on.
10. mediator's dough sheet collection of having survived, and need not generate from volume data again.
returns main interface or withdraws from.
The implement device data volume is big; Unstructured information; Uncertain information; Continuous media such as sound, video has the time sensitivity characteristics; Because the characteristics of multimedia messages integration cause that relation is complicated between the different medium; The hardware that multimedia messages is handled has correlativity:
(1) support figure, image, animation, sound, dynamically advise multimedia field type and user definition specific type such as hungry text;
(2) integrated management of support fixed-length data and non-fixed-length data;
(3) expression and the processing of the complicated entity of support, requirement has the ability of complex relationship (as time-space relationship) between expression and processing entities, guaranteed complicated entity integrity and conforming mechanism;
(4) support the multiple form of expression (when playing, can change its frequency or a width of cloth still image, when showing, change character such as its contrast and do not influence content in the storehouse) of same entity as one section video;
(5) has good user interface;
(6) support multimedia special inquiry and good Processing Interface;
(7) support distributed environment.
Based on the new and high technology market defining standard and the implement device thereof of grid computing, digital signal processing is exactly that conversion is converted to output signal the various computings of a series of input signals process, and how quick its key just is, carries out a large amount of mathematical operations efficiently.Common computer all has the function of data processing and mathematical operation, take into account and all reach optimal effectiveness but be difficult to both, and digital signal processor is a class microprocessor that designs at mathematical operation special in the digital signal processing.The bottleneck of a maximum of combine digital signal processing algorithm is a frequent statement mass data information from storer, comprises the input signal sampling, filter coefficient and programmed instruction or the like.At this problem, digital signal processor adopted Harvard (Harvard) structure with program and data storage in different storage spaces, be that program storage and data storage are two separate storeies, each storer independently addresses, independent access, the access space that is programmed instruction and data separates, and program bus and the data bus of oneself respectively arranged.Simultaneously deal with data and program of processor like this, the earth ground has improved the processing power of processor.
The one, allow deposit data in program storage, and directly used by arithmetic operation instruction, strengthened the dirigibility of chip; The 2nd, instruction is stored in the high-speed buffer (Cache), when carrying out this instruction, does not need reading command from chip external memory again, has saved the time of an instruction cycle.
Dsp chip extensively adopts streamline reducing instruction time, thereby has strengthened the processing power of processor.
Multitask is provided, multithreading, micronucleus is simplified the exploitation under the multi-processor environment in real time.Micronucleus is write as with C language and assembly language, takies the space of 3Kwords (32bit).All move the copy of a micronucleus in the system on each DSP, be responsible for task, the environmental protection that the management of thread and external interrupt cause, memory management or the like by it.
Provide " virtual communication port " driver as the standard device of the communication between processor, and with the hardware independent of system, no matter you adopt shared drive or serial ports interconnection.Abundant C built-in function uses these function calls " virtual communication port " driver, can realize the communication and synchronously between processor easily.
Based on the ARM of the embedded computing technique in the world (have unique RISC technology and be better than present many competitive structures, logic advanced person, diverse in function, low consumption at a low price, simplicity of design is easy to integrated.) dsp chip (SOC) solution that provides is numerous in variety, all makes the basic structure of microprocessor, peripherals and System on Chip/SoC as exploitation with embedded computing technique.
The embedded computing technique IP in the world of ARM (semiconductor intellecture property supplier) provides and comprises chip core, structure extension, SDK (Software Development Kit) and System on Chip/SoC solution.The exploitation and the support environment of these technology are first-class, meet the requirement of the commerce and the consumer applications of current fast development, have become the basis of the digital electronic product that can improve constantly.MIPS/ watt also quite superior with code density.When the former code chip of software solidified, the real-time transparent decompress(ion) of 32 bit instructions was 64, does not lose any performance.Therefore software design can be used ARM code and Thumb (instruction set is the expansion of ARM code structure) code in the digital applications system, reached maximum application flexibility.ARM code and Thumb instruction set, but S synthesis processor series is to strengthen digital signal processor (DSP), comprises ARM9E-STM and ARM966E-
STM cache processor macroelement, suitable application system of mixing DSP and microcontroller performance.Single-chip ARM966E-
S series is greatly dwindled chip volume, complicacy and power consumption, and can shorten the listing cycle.The kernel of employing Thumb instruction set technology comprises the digital processing expansion of ARM instruction set, obtains comprising a large amount of third party technology supports such as operating system, real time operating system, application software, SDK (Software Development Kit), electric design automation (EDA) instrument, development template.ARM966E-S comprises that EmbeddedICE JTAG software debugging aids enhanced edition---EmbeddedICE-RTTM logic more meets the requirement of real-time system development.ARM966E-S series is widely used in intelligent digital player, digitizing mass memory unit, speech coder, PDA, smart phone, Video Codec, MPEG series, MP3 audio decoder, speech recognition and synthetic, and the automatic control solution that comprises Ruiss Controll and little processing.
ARM10 and Thumb instruction set are 400+MIPS processor of new generation, aim at efficient handheld digital device design such as digital processing system and manager and smart phone.This code command energy collecting enough reduces system complexity, increase dirigibility, and for various digital processing systems provide the processor macroelement of low-cost and high-performance, and for complex video visual field boundary scape operation such as interactive and man-machine interaction is exported user interface senior integer and floating-point operation ability is provided.
ARM and Thumb instruction set aim at the AMBATM bus compatible IP kernel of the integrated design of chip, and the AMBA on-chip bus is set open standard, are the structures of System on Chip/SoC design, also are that IP utilizes and the key that effectively is connected " the digital glue " of IP kernel again.AMBA is becoming the standard of system architecture and the exploitation of IP storehouse rapidly.
By the method for configuration software, allocating task between processor easily, optimization system performance.The input-output function of support standard is conveniently transferred survey.
The maximum characteristics of this product are the DSP structures of SIMD (single-instruction multiple-data stream (SIMD)) pattern, and this texture ratio first generation SHARC structure has increased by the second cover computing unit, and have improved the data locomotivity of processor greatly.Under two processing elements (each processing elements comprises multiplier, ALU, shift unit and register file etc.) structure, instruction can be in two processing units executed in parallel, thereby realize parallel processing multiple data stream (SIMD pattern).Make computing velocity double.In order to realize parallel processing between a plurality of DSP sheets better, ADSP21160 has processor interface and supports shared total county of multiprocessor, the internal data bus width is brought up to 64, the DMA passage is increased to 14 from 10, because storer is a twoport, DMA can with core processor access memory simultaneously, do not clash, do not insert additional cycles.

Claims (1)

1, a kind of implement device of the new and high technology market defining standard based on grid computing comprises: target-oriented market defining virtual and real boundary, the grid computing cooperative working environment, the multi-channel data source, attributive character, pattern-recognition, component identification, sight is set, carry out non-linear task, the man-machine interaction interpreter, evil spirit is carried out the coffin upon burial and is held up, visual field boundary scape operation, above its module, comprise: object identification monitoring and control configuration module, the mathematical model of market defining standard, non-process template, picture of world, intelligence gateway module, below this module, comprising: the artificial intelligence network of grid computing, the evil spirit of visual field boundary scape operation is carried out the coffin upon burial and is held up, the real-time information storehouse, the object resources bank, association's knowledge base, the case analysis storehouse.
CNA031169112A 2003-05-15 2003-05-15 Apparatus for realizing high-new technology market fixed standard based on net computation Pending CN1549171A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CNA031169112A CN1549171A (en) 2003-05-15 2003-05-15 Apparatus for realizing high-new technology market fixed standard based on net computation

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CNA031169112A CN1549171A (en) 2003-05-15 2003-05-15 Apparatus for realizing high-new technology market fixed standard based on net computation

Publications (1)

Publication Number Publication Date
CN1549171A true CN1549171A (en) 2004-11-24

Family

ID=34320528

Family Applications (1)

Application Number Title Priority Date Filing Date
CNA031169112A Pending CN1549171A (en) 2003-05-15 2003-05-15 Apparatus for realizing high-new technology market fixed standard based on net computation

Country Status (1)

Country Link
CN (1) CN1549171A (en)

Cited By (20)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN100498833C (en) * 2005-09-30 2009-06-10 中国科学院计算机网络信息中心 Computing grid long-distance visualization method
CN102037490A (en) * 2008-09-25 2011-04-27 电子地图有限公司 Method of and arrangement for blurring an image
CN102446226A (en) * 2012-01-16 2012-05-09 上海方正数字出版技术有限公司 Method for achieving NoSQL key-value storage engine
CN103377432A (en) * 2012-04-16 2013-10-30 殷程 Intelligent customer service marketing analysis system
CN103617201A (en) * 2013-11-14 2014-03-05 陕西理工学院 Visualization method and visualization model
CN106709507A (en) * 2016-11-29 2017-05-24 北京林业大学 Force-directed segmented bone parallel coordinates plot clustering data bundling method
CN106997414A (en) * 2017-04-06 2017-08-01 北京仿真中心 A kind of multidisciplinary collaboration Simulation Methods based on reflective memory network
CN107194490A (en) * 2016-03-14 2017-09-22 商业对象软件有限公司 Predict modeling optimization
CN108171333A (en) * 2018-01-22 2018-06-15 南京邮电大学 Artificial intelligence body rescue Communication Decision optimization method under low communication environment
CN108287876A (en) * 2017-12-29 2018-07-17 国网北京市电力公司 Support the power quality data service quality detection method and device of multiple format
CN109214620A (en) * 2017-06-29 2019-01-15 格局商学教育科技(深圳)有限公司 The creation method and interprise's intensive management plateform system of interprise's intensive management platform
TWI656455B (en) * 2017-05-18 2019-04-11 華南商業銀行股份有限公司 Server for providing mixed reality environment
CN110287879A (en) * 2019-06-26 2019-09-27 天津大学 A kind of video behavior recognition methods based on attention mechanism
CN110363382A (en) * 2019-06-03 2019-10-22 华东电力试验研究院有限公司 Almightiness type Township Merging integrated business integration technology
CN112639888A (en) * 2018-08-09 2021-04-09 祖克斯有限公司 Programmed world generation
CN114547380A (en) * 2022-01-25 2022-05-27 北京元年科技股份有限公司 Data traversal query method and device, electronic equipment and readable storage medium
CN114662925A (en) * 2022-03-25 2022-06-24 国网河北省电力有限公司 Data fusion and mining method for power service center
CN115049799A (en) * 2022-06-14 2022-09-13 北京百度网讯科技有限公司 Method and device for generating 3D model and virtual image
CN115470831A (en) * 2022-11-03 2022-12-13 四川中电启明星信息技术有限公司 Hydropower signal abnormity judgment method based on frequent item set reasoning
CN116820772A (en) * 2023-07-03 2023-09-29 中山大学 Grid parallel reading method, device, terminal equipment and readable storage medium

Cited By (27)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN100498833C (en) * 2005-09-30 2009-06-10 中国科学院计算机网络信息中心 Computing grid long-distance visualization method
CN102037490A (en) * 2008-09-25 2011-04-27 电子地图有限公司 Method of and arrangement for blurring an image
CN102446226A (en) * 2012-01-16 2012-05-09 上海方正数字出版技术有限公司 Method for achieving NoSQL key-value storage engine
CN102446226B (en) * 2012-01-16 2015-09-16 北大方正集团有限公司 A kind of method realizing the key assignments storage engines of NoSQL
CN103377432A (en) * 2012-04-16 2013-10-30 殷程 Intelligent customer service marketing analysis system
CN103617201A (en) * 2013-11-14 2014-03-05 陕西理工学院 Visualization method and visualization model
CN107194490A (en) * 2016-03-14 2017-09-22 商业对象软件有限公司 Predict modeling optimization
CN107194490B (en) * 2016-03-14 2022-08-12 商业对象软件有限公司 Predictive modeling optimization
CN106709507A (en) * 2016-11-29 2017-05-24 北京林业大学 Force-directed segmented bone parallel coordinates plot clustering data bundling method
CN106709507B (en) * 2016-11-29 2019-11-08 北京林业大学 A kind of parallel coordinate system view cluster data binding method of power guiding segmentation bone
CN106997414A (en) * 2017-04-06 2017-08-01 北京仿真中心 A kind of multidisciplinary collaboration Simulation Methods based on reflective memory network
CN106997414B (en) * 2017-04-06 2020-08-14 北京仿真中心 Multidisciplinary collaborative simulation implementation method based on reflective memory network
TWI656455B (en) * 2017-05-18 2019-04-11 華南商業銀行股份有限公司 Server for providing mixed reality environment
CN109214620A (en) * 2017-06-29 2019-01-15 格局商学教育科技(深圳)有限公司 The creation method and interprise's intensive management plateform system of interprise's intensive management platform
CN108287876A (en) * 2017-12-29 2018-07-17 国网北京市电力公司 Support the power quality data service quality detection method and device of multiple format
CN108171333A (en) * 2018-01-22 2018-06-15 南京邮电大学 Artificial intelligence body rescue Communication Decision optimization method under low communication environment
CN112639888A (en) * 2018-08-09 2021-04-09 祖克斯有限公司 Programmed world generation
CN110363382A (en) * 2019-06-03 2019-10-22 华东电力试验研究院有限公司 Almightiness type Township Merging integrated business integration technology
CN110287879B (en) * 2019-06-26 2023-01-17 天津大学 Attention mechanism-based video behavior identification method
CN110287879A (en) * 2019-06-26 2019-09-27 天津大学 A kind of video behavior recognition methods based on attention mechanism
CN114547380A (en) * 2022-01-25 2022-05-27 北京元年科技股份有限公司 Data traversal query method and device, electronic equipment and readable storage medium
CN114662925A (en) * 2022-03-25 2022-06-24 国网河北省电力有限公司 Data fusion and mining method for power service center
CN115049799A (en) * 2022-06-14 2022-09-13 北京百度网讯科技有限公司 Method and device for generating 3D model and virtual image
CN115049799B (en) * 2022-06-14 2024-01-09 北京百度网讯科技有限公司 Method and device for generating 3D model and virtual image
CN115470831A (en) * 2022-11-03 2022-12-13 四川中电启明星信息技术有限公司 Hydropower signal abnormity judgment method based on frequent item set reasoning
CN116820772A (en) * 2023-07-03 2023-09-29 中山大学 Grid parallel reading method, device, terminal equipment and readable storage medium
CN116820772B (en) * 2023-07-03 2024-05-28 中山大学 Grid parallel reading method, device, terminal equipment and readable storage medium

Similar Documents

Publication Publication Date Title
CN1549171A (en) Apparatus for realizing high-new technology market fixed standard based on net computation
Bhatia Data mining and data warehousing: principles and practical techniques
CN1549182A (en) Arrangement for realizing new technique
Giorgini et al. GRAnD: A goal-oriented approach to requirement analysis in data warehouses
Mandapuram et al. Investigating the Prospects of Generative Artificial Intelligence
Tarnowska et al. Recommender system for improving customer loyalty
English Business intelligence as a source of competitive advantage in SMEs: A systematic review
Wu et al. Sustainable development performance for small and medium enterprises using a fuzzy synthetic method-DEMATEL
Tomar et al. The role of reference data in financial data analysis: Challenges and opportunities
Akerkar Advanced data analytics for business
Pisoni et al. Data science for finance: Best-suited methods and enterprise architectures
Martínez et al. An RFM model customizable to product catalogues and marketing criteria using fuzzy linguistic models: Case study of a retail business
Skersys et al. The enrichment of BPMN business process model with SBVR business vocabulary and rules
Zhang et al. Developing evaluation frameworks for business models in China’s rural markets
Fu et al. User-Driven: A Product Innovation Design Method for a Digital Twin Combined with Flow Function Analysis
Ha et al. Job forecasting based on the patent information: a word embedding-based approach
Cheong et al. A Data Valuation Model to Estimate the Investment Value of Platform Companies: Based on Discounted Cash Flow
Adams Perspectives on data mining
Park et al. Spec guidance for engineering design based on data mining and neural networks
Peng et al. Integration of deep reinforcement learning with collaborative filtering for movie recommendation systems
Wang et al. An Integrated Data-Driven Procedure for Product Specification Recommendation Optimization with LDA-LightGBM and QFD
Rofi'i Analysis of E-Commerce Purchase Patterns Using Big Data: An Integrative Approach to Understanding Consumer Behavior
JP5018313B2 (en) Information structuring apparatus, information structuring method, and program
ul Hasan et al. Investigating consumer values of secondhand fashion consumption in the mass market vs. luxury market: a text-mining approach
Evelson The Forrester Wave™: Enterprise Business Intelligence Platforms, Q4 2010

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C02 Deemed withdrawal of patent application after publication (patent law 2001)
WD01 Invention patent application deemed withdrawn after publication