CN101615265B - Intelligent decision simulating experimental system based on multi-Agent technology - Google Patents
Intelligent decision simulating experimental system based on multi-Agent technology Download PDFInfo
- Publication number
- CN101615265B CN101615265B CN2009100130451A CN200910013045A CN101615265B CN 101615265 B CN101615265 B CN 101615265B CN 2009100130451 A CN2009100130451 A CN 2009100130451A CN 200910013045 A CN200910013045 A CN 200910013045A CN 101615265 B CN101615265 B CN 101615265B
- Authority
- CN
- China
- Prior art keywords
- agent
- decision
- module
- planning
- dictionary
- 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.)
- Expired - Fee Related
Links
- 238000005516 engineering process Methods 0.000 title claims abstract description 26
- 238000000034 method Methods 0.000 claims abstract description 117
- 238000013439 planning Methods 0.000 claims abstract description 105
- 230000006870 function Effects 0.000 claims abstract description 77
- 238000007726 management method Methods 0.000 claims abstract description 77
- 238000004088 simulation Methods 0.000 claims abstract description 49
- 238000002474 experimental method Methods 0.000 claims abstract description 38
- 238000013473 artificial intelligence Methods 0.000 claims abstract description 11
- 238000012545 processing Methods 0.000 claims abstract description 10
- 238000004891 communication Methods 0.000 claims description 125
- 239000010410 layer Substances 0.000 claims description 114
- 238000004422 calculation algorithm Methods 0.000 claims description 100
- 230000008569 process Effects 0.000 claims description 98
- 239000000463 material Substances 0.000 claims description 66
- 230000006399 behavior Effects 0.000 claims description 60
- 238000004458 analytical method Methods 0.000 claims description 32
- 230000000875 corresponding effect Effects 0.000 claims description 29
- 230000004044 response Effects 0.000 claims description 24
- 238000003058 natural language processing Methods 0.000 claims description 23
- 238000011156 evaluation Methods 0.000 claims description 21
- 238000003860 storage Methods 0.000 claims description 14
- 230000009471 action Effects 0.000 claims description 12
- 230000007246 mechanism Effects 0.000 claims description 11
- 238000013178 mathematical model Methods 0.000 claims description 10
- 230000000007 visual effect Effects 0.000 claims description 9
- 238000007418 data mining Methods 0.000 claims description 7
- 230000015572 biosynthetic process Effects 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 6
- 238000001914 filtration Methods 0.000 claims description 6
- 238000012423 maintenance Methods 0.000 claims description 6
- 230000008878 coupling Effects 0.000 claims description 5
- 238000010168 coupling process Methods 0.000 claims description 5
- 238000005859 coupling reaction Methods 0.000 claims description 5
- 238000000354 decomposition reaction Methods 0.000 claims description 5
- 238000009795 derivation Methods 0.000 claims description 5
- 238000009826 distribution Methods 0.000 claims description 5
- 230000006872 improvement Effects 0.000 claims description 5
- 230000005540 biological transmission Effects 0.000 claims description 4
- 230000007613 environmental effect Effects 0.000 claims description 4
- 238000012360 testing method Methods 0.000 claims description 4
- 238000010835 comparative analysis Methods 0.000 claims description 3
- 230000002860 competitive effect Effects 0.000 claims description 3
- 230000008676 import Effects 0.000 claims description 3
- 238000013459 approach Methods 0.000 claims description 2
- 230000008859 change Effects 0.000 claims description 2
- 238000011157 data evaluation Methods 0.000 claims description 2
- 238000013523 data management Methods 0.000 claims description 2
- 238000013524 data verification Methods 0.000 claims description 2
- 238000000605 extraction Methods 0.000 claims description 2
- 239000002346 layers by function Substances 0.000 claims description 2
- 238000012986 modification Methods 0.000 claims description 2
- 230000004048 modification Effects 0.000 claims description 2
- 230000008447 perception Effects 0.000 claims description 2
- 238000012800 visualization Methods 0.000 claims 1
- 230000000694 effects Effects 0.000 abstract description 4
- 238000005094 computer simulation Methods 0.000 abstract description 3
- 238000011161 development Methods 0.000 abstract description 3
- 238000013461 design Methods 0.000 description 64
- 238000010586 diagram Methods 0.000 description 8
- 238000009412 basement excavation Methods 0.000 description 6
- 230000003993 interaction Effects 0.000 description 6
- 239000000284 extract Substances 0.000 description 5
- TVZRAEYQIKYCPH-UHFFFAOYSA-N 3-(trimethylsilyl)propane-1-sulfonic acid Chemical compound C[Si](C)(C)CCCS(O)(=O)=O TVZRAEYQIKYCPH-UHFFFAOYSA-N 0.000 description 4
- 238000012937 correction Methods 0.000 description 3
- 238000011160 research Methods 0.000 description 3
- 230000018109 developmental process Effects 0.000 description 2
- 238000012854 evaluation process Methods 0.000 description 2
- 239000000203 mixture Substances 0.000 description 2
- 230000008520 organization Effects 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 230000032683 aging Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 239000012141 concentrate Substances 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 230000002950 deficient Effects 0.000 description 1
- 238000000151 deposition Methods 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 230000000977 initiatory effect Effects 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 238000012827 research and development Methods 0.000 description 1
- 238000004904 shortening Methods 0.000 description 1
Images
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to an intelligent decision simulating experimental system based on multi-Agent technology, belonging to the crossed field of computer simulation, artificial intelligence and management experiment, and being the breakthrough and the development of a management experimental methodology and an intelligent decision support system on the concept, the theory, the technology and the function. The system is a large-scale experimental platform for carrying out simulation assisted prediction and decision on discrete events in management activities of the real world. In the method, an isomerized Agent and the group collaboration thereof are creatively used as leading roles of transparent processing of a decision experiment and a system simulation, four functional sub-systems, four layers of Agent system structures and four types of local databases are established, and predictions and decisions of four simulating experimental methods are applied to unify a system module. The invention is based on an internet, synthetically uses several new technologies of cloud computing and the artificial intelligence, increases functions of games of strategy, machine planning, knowledge discovery and logic self-building, and has super experimental capability. The invention is a new way for developing the management experimental technology and the intelligent decision support system.
Description
Technical field
The present invention relates to a kind of intelligence simulation system of management experiment or the decision support based on multi-Agent technology.Its notion may be defined as: intelligent decision simulating experimental system (Intelligent Decision Simulation Experimental System; IDSES); The crossing domain that belongs to Computer Simulation, artificial intelligence and management experiment is the more advanced intelligent decision system of multinomial new technology, cloud computing and the management experimental methodology of a kind of integrated system emulation, computer graphics, artificial intelligence.
Background technology
Preparedness ensures success, unpreparedness spells failure, and No preparation, no success, and the management experiment has important role for organizational decision making, reduces cost, predicts the outcome and avoid risk like optimal design, raises the efficiency the shortening cycle, and virtual reality reduces material consumption.Be that operational level or the decision-making of strategic level (Scheme Choice) all are the higher management of tissue, and the science decision of low cost, low-risk, high-level efficiency, high safety become the demand emphasis of all kinds of organization and administration practices already.Research shows that China is more scattered and rare to the research of management experiment at present, and mainly concentrates on the computer based simulated experiment, and most studies is limited in the running research, and wherein the method for system dynamics is main simulation means.
Based on the emulation technology of computing machine and network thereof, not only can raise the efficiency, shorten the research and development cycle, reduce the training time, not limited by environment and weather, and to guarantee safely, cut down expenses, improving the quality has outstanding effect.Emulation technology be build on experimental conceptive.When wanting the decision of mechanism to use a new design or new when tactful, often because the restriction of time and fund has no idea to bear the risk that failure brings, emulation technology can help it to alleviate the risk of failure at this moment.Through the situation of computer virtual reality, the decision maker can know the feasibility of strategy or design, thereby helps them to make wise decision-making.Developed country such as the U.S., Japan is leading aspect Computer Simulation; Successively developed like instruments such as WINTNES, FLEXSIM, RaLC; But these emulation tools all are the masters that is applied as with work layers such as determinacy planning in kind and process simulations, lack the emulation of strategic decision emulation and dynamic uncertainty environment.The main emulation mode of these instruments has mathematical simulation, vector graphics, three-dimensional model etc.They possess certain flexibility and visual, but intellectuality and integrated degree are low, and function is simple.Along with the object scale that the mankind studied is huge day by day, structure is complicated day by day, and environment is uncertain day by day, and organizational decision making only relies on people's experience and traditional technology to be difficult to satisfy increasingly high requirement.
Current intelligent decision support system (IDSS) notion is cut people such as gram the earliest and is proposed the eighties in 20th century by American scholar Bonn, its function is, can handle quantitative problem, can handle qualitative question again.Be that artificial intelligence (AI) and DSS combine; The application of expert system technology; Make DSS can use human knowledge more fully, like descriptive knowledge about decision problem, the procedural knowledge in the decision process; Find the solution the reasoning sex knowledge of problem, help solve the aid decision-making system of complicated decision problem through reasoning from logic.The core concept of IDSS is that AI is combined with other related science achievement, makes DSS have artificial intelligence.But the maximum defective of IDSS is to lack the emulation experiment process, i.e. forecasting process, especially visual simulating process, and the effect of aid decision making can't precheck.Efficient can become very low when running into the expertise shortage and needing the large-scale complex parallel computation in addition.
Yet the competition of modern tissue is lean more and more, and an error just possibly forfeited chance.Therefore intelligentized prediction integrated experiment porch of making a strategic decision seems most important for science, the effective decision-making of tissue.The intelligent decision simulating experimental system integrated use the multinomial new technology of cloud computing and artificial intelligence; Like many agent technology, natural language processing, machine learning, Knowledge Discovery etc.; Than traditional simulation system and more intellectuality of DSS, openingization, integrated and efficient, function is more comprehensive.
Summary of the invention
The objective of the invention is in order to make full use of the state-of-the-art technology of artificial intelligence; In conjunction with the predictable of emulation technology the management experiment is broken through with technical in theory; The design concept and the implementation method of innovation intelligent decision support system; On theory, method, technology and function, form the new ideas of existing relatively IDSS, form the Predicting and Policy-Making integrative intelligent system, all include tissue in experimental system from work layer to the decision-making of strategy layer; With intelligent visual emulation experiment forecasting process and result, reduce organizational decision making's cost, evade risk of policy making, improve decision science property, validity and efficient.Intelligent decision simulating experimental system is an extensive decision experiments platform; Open and flexible fully; The user can set up logical base, knowledge base, material database, the example/countermeasure storehouse of oneself on the basis of this platform; Utilize the powerful experiential function (3D model emulation, 2D graphical simulation, mathematical model simulation, reasoning from logic emulation and cloud computing) of platform to set up the laboratory of own industry.This system can be applicable to fields such as enterprise management decision-making, governability decision-making, command of armed force decision-making, education of universities and colleges training.
The present invention addresses the above problem the technical scheme that is adopted: a kind of intelligent decision simulating experimental system (IDSES) based on multi-Agent technology, it be one based on internet, integrated Predicting and Policy-Making integral intelligent analogue system;
(1) said intelligent decision simulating experimental system comprises planning subsystem, evaluation subsystem, decision-making subsystem, four functional subsystems of game subsystem;
(2) said each functional subsystem has four layers of Agent system structure (MAS), promptly functional layer MAS, work layer MAS, communication layers MAS, resource layer MAS;
(3) said planning subsystem, its function are to utilize this emulation experiment platform to carry out intelligence machine planning or artificial planning proposition decision scheme; Function layer MAS comprises: man-machine interface Agent, explanation Agent, manual intervention Agent, constraint Agent, reasoning Agent; , work layer MAS comprises: planning Agent, task Agent, project Agent, operation Agent, Information Agent, natural language processing Agent, emulation Agent, graphic modeling Agent; Communication layers MAS comprises: communication Agent; Resource layer MAS comprises: material database Management Agent, KBM Agent, logical base Management Agent, example/countermeasure library management Agent, teledata are excavated Agent;
(4) said evaluation subsystem, its function are to utilize this emulation experiment platform that the decision-making prediction scheme is carried out intelligent evaluation to suggest improvements; Function layer MAS comprises: man-machine interface Agent, explanation Agent, manual intervention Agent; Work layer MAS comprises: assessment Agent, natural language processing Agent, emulation Agent, graphic modeling Agent; Communication layers MAS comprises: communication Agent; Resource layer MAS comprises: material database Management Agent, KBM Agent, logical base Management Agent, example/countermeasure library management Agent, teledata are excavated Agent;
(5) said decision-making subsystem, its function are to utilize this emulation experiment platform that a plurality of decision-making prediction schemes are carried out intellectual analysis to select satisfactory solution; Function layer MAS comprises: man-machine interface Agent, explanation Agent, manual intervention Agent; Work layer MAS comprises: decision analysis Agent, natural language processing Agent, emulation Agent, graphic modeling Agent; Communication layers MAS comprises: communication Agent; Resource layer MAS comprises: material database Management Agent, KBM Agent, logical base Management Agent, example/countermeasure library management Agent, teledata are excavated Agent;
(6) said game subsystem, its function are to utilize this emulation experiment platform to carry out intelligent man-machine game or the intelligent online game competitive power with the check decision-making; Function layer MAS comprises: man-machine interface Agent, explanation Agent, manual intervention Agent; Work layer MAS comprises: game Agent, tactful Agent, project Agent, operation Agent, Information Agent, target Agent, natural language processing Agent, emulation Agent, graphic modeling Agent; Communication layers MAS comprises: communication Agent; Resource layer MAS comprises: material database Management Agent, KBM Agent, logical base Management Agent, countermeasure library management Agent, teledata are excavated Agent;
(7) said each subsystem function layer MAS, functional in order to the fulfillment upper strata:
Man-machine interface Agent: response type, be used for obtaining information, to the user conclusion is provided, the request user imports necessary information, comprises input/output control module, communication module, behavior sensing module, I/O algorithm data-base and Agent state table and Agent dictionary;
Explain Agent: mixed type; Be used for experimental result is carried out necessary, rational explanation to operational process; Feed back to man-machine interface Agent to the result and export to the user, comprise and explain maker, conclusion control module, communication module, behavior sensing module, interpretation algorithms database and Agent state table and Agent dictionary;
Manual intervention Agent: response type, be used for when inadequate resource, starting corresponding program, handle by user intervention, comprise routine call control module, communication module, behavior sensing module, call algorithm data-base;
Constraint Agent: response type, be used to solve the constraint satisfaction of decision experiments process, the constraint that mainly solves common-sense and agreement comprises bound control module, communication module, behavior sensing module, bounding algorithm database and constraint condition table and Agent dictionary;
Reasoning Agent: i.e. inference machine; Mixed type; Rely on expertise and strategy that the machine planning scheme is carried out true inference; Agent suggests improvements to planning, comprises problem deduce machine, regular control module, communication module, behavior sensing module, reasoning algorithm database and rule set and Agent dictionary;
(8) said each subsystem work layer MAS, in order to realize system core business operation function:
Planning Agent: mixed type; Be used for accepting user's planning experiment request; Set up planning constraint, plan model and planning territory; The program results of integrated each task Agent forms the general planning case, comprises mission planning device, decision-making and intelligent control module, communication module, behavior sensing module, planning algorithm database and Agent state table and Agent dictionary;
Assessment Agent: mixed type; Be used for accepting user's assessment experiment request; Set up evaluation index system and test set; The target that the decision scheme user is concerned about compares evaluation and reasoning from logic evaluation, forms evaluation of programme, comprises evaluate alternatives module, assessment control module, communication module, behavior sensing module, assessment algorithm database and Agent state table and Agent dictionary;
Decision analysis Agent; Mixed type; Be used for accepting user's decision analysis experiment request; Set up decision-making constraint, object module and choice mechanism; All kinds of schemes that the comparative analysis user submits to form the decision-making suggestion, comprise scheme comparison module, decision analysis control module, communication module, behavior sensing module, decision making algorithm database and Agent state table and Agent dictionary;
Game Agent: mixed type; Be used for accepting user's game experiment request; Set up judge's model, inference mechanism and game constraint; The various schemes that the ruling user submits to form payoff, comprise strategy generator, ruling and intelligent control module, communication module, behavior sensing module, game playing algorithm database and Agent state table and Agent dictionary;
Task Agent: mixed type; Carry out PROBLEM DECOMPOSITION according to projects Agent ability and state; Give corresponding project Agent Task Distribution; The sub-case of planning of assembled item Agent feeds back to planning Agent simultaneously, comprises project planning device, decision-making and intelligent control module, communication module, behavior sensing module, planning algorithm database and Agent state table and Agent dictionary;
Target Agent: response type, be used on network, seeking online game target and strategy thereof and feed back to fight and play chess Agent, comprise target search control module, communication module, behavior sensing module, searching algorithm database and routing table and Agent dictionary;
Strategy Agent: mixed type; Effect is with task Agent; Be controlled different with the transmission object, result formats is different, comprises project planning device, decision-making and intelligent control module, communication module, behavior sensing module, planning algorithm database and Agent state table and Agent dictionary;
Project Agent: mixed type; A plurality of; It is the main body of accomplishing experimental duties; Bear concrete experimental duties, can derive down one deck operation Agent, comprise work planning device, decision-making and intelligent control module, communication module, behavior sensing module, planning algorithm database and Agent state table and Agent dictionary;
Operation Agent: mixed type; A plurality of; Be the base unit of accomplishing experimental duties,, utilize relevant information formation base experimental result and feed back to project Agent according to resulting subproblem and parameter; Comprise the subproblem solver, find the solution and intelligent control module, communication module, behavior sensing module, derivation algorithm database and Agent state table and Agent word;
Information Agent: response type; A plurality of; Be responsible for collecting required various information of upper level Agent and feedback; Possibly need a plurality of Information Agent to accomplish various tasks and to hold consultation and cooperate, comprise information acquisition control module, communication module, behavior sensing module, gather algorithm data-base and Agent state table and Agent dictionary;
Natural language processing Agent: mixed type; Be used for process user at experimental system interface input and extensive real text system's output; Carry out information retrieval filtration and information extraction; The succinct authority data that converts system identification to is handled to treat core Agent, comprises language processing module, information Recognition and intelligent control module, communication module, behavior sensing module, recognizer database, corpus and proprietary vocabulary and Agent dictionary;
Emulation Agent: response type, be used for the standardization decision scheme is carried out mathematical simulation, calculating and reasoning from logic emulation and exports the result, comprise Simulation Control module, communication module, behavior sensing module, simulation algorithm database and function table and Agent dictionary;
Graphic modeling Agent: response type; Be used for simulation process with suitable figure (two dimension, three-dimensional) visual Simulation; Have spatial parameter and behavior calculation of parameter function simultaneously, comprise graphic modeling control module, communication module, behavior sensing module, action algorithm data-base and function table and Agent dictionary;
(9) said each subsystem communication layer MAS has only communication Agent, response type; In order to realize data queue; The registration of Agent is provided, message based communication, the supplier provides coupling for message; Functions such as interpretation process comprise the communication server, message transmission module, behavior sensing module, communication of algorithms database and registration table and routing table;
(10) said each subsystem resource layer MAS is in order to realize the system bottom data management function;
The material database Management Agent: response type, be responsible for material database management maintenance, material combination and operation, comprise the material classification and call control module, communication module, behavior sensing module, classify and call algorithm data-base with Agent state table and Agent dictionary;
KBM Agent: mixed type; Be responsible for MAINTENANCE OF KNOWLEDGE BASE, knowledge is relatively upgraded and knowledge-based inference work, comprise knowledge comparison update module, knowledge classification and call control module, knowledge-based inference module, communication module, behavior sensing module, comparison algorithm database, classify and call algorithm data-base and Agent state table and Agent dictionary;
Logical base Management Agent: mixed type; Maintenance, the logic of being responsible for logical base called, logic is mated and the reasoning work of logic-based, comprise logic comparison update module, logical division and call control module, logic-based reasoning module, communication module, behavior sensing module, comparison algorithm database, classify and call algorithm data-base and Agent state table and Agent dictionary;
Example/countermeasure library management Agent: mixed type; Be responsible for the retrieval, coupling, modification in example/countermeasure storehouse, the corresponding function based on case-based reasoning such as multiplexing, comprise example/countermeasure comparison update module, example/classification of game and call control module, communication module, behavior sensing module, comparison algorithm database, classify and call algorithm data-base and Agent state table and Agent dictionary;
Teledata is excavated Agent: mixed type; Be responsible for carrying out when local data lacks that teledata is obtained, data are tested and data collision and redundancy are cleared up, comprise that data evaluation and verification module, data mining control module, communication module, behavior sensing module, KDD component base and Internet resources retrieve and the Agent dictionary;
(11) said response type Agent: make the respective reaction action through the perception environmental change, represent Agent::=< Aid, P, A, see, action>with five-tuple, wherein, Aid is the sign of certain concrete Agent; A representes Agent behavior collection; P representes Agent visual state collection; See, action are used to portray inner observation process of Agent and behaviour decision making process;
(12) said mixed type Agent: structure is divided into two-layer, and bottom is a responding layer, and symbolization is not represented and reasoning, can respond and handle the sudden variation of external environment condition fast, has higher priority usually; High-rise adopt that traditional artificial intelligence approach is planned, reasoning and decision-making, represent Agent::=< Aid, P, A with 11 tuples; R, Dva, Dar; Rule, see next, estimate; Action >, wherein, R representes the set that all corresponding possible outcome states of all possible action schemes of Agent are constituted; Dva representes the rule of correspondence collection of visual state and scheme set; Corresponding relation between Dar representation scheme collection and the result phase; Rule is the set that concrete decision rule constitutes; Next is used to portray the inner thought process of Agent; Estimate representes selected scheme set, according to the rule of correspondence storehouse of scheme and result phase collection, confirms possible result phase.Other element implication is identical with response type Agent.
(13) said intelligent decision simulating experimental system comprises four types of local data bases such as material database, example/countermeasure storehouse, knowledge base, logical base; Material database is in order to storage and provide graphic modeling needed figure material; Knowledge base is in order to storage and the needed expertise of experiment, experimental rules, managerial knowledge and information data are provided; Logical base is in order to storage and the needed operation logic of experiment, inference logic, interpretation logic etc. are provided; Example/countermeasure storehouse is in order to storage and Success in Experiment example/countermeasure and the needed same or similar case/countermeasure of experiment are provided.
(14) said intelligent decision simulating experimental system utilization mathematical model simulation, X-Y scheme emulation, three-dimensional model emulation, 4 kinds of emulation experiment methods of reasoning from logic emulation; Mathematical model simulation is abstract and simplification with problem, sets up mathematical model, carries out behavior emulation through solution procedure, thereby predicts the outcome; Mathematical model simulation can combine with graphical simulation, reasoning from logic; X-Y scheme emulation is to utilize machine or artificial foundation, with dynamic planar graph emulation expression behaviour process, thus observation process and result of calculation; Three-dimensional model emulation is to utilize machine or artificial foundation, with the comprehensive expression behaviour process of dynamic 3 D graphics emulation, thus observation process and result of calculation; Reasoning from logic emulation is to give expertise with problem and environmental parameter, utilizes inference mechanism and algorithm predicts process and result, combines with mathematical model simulation usually.
Meaning of the present invention is: 1) be management experimental methodology, system simulation technology and integration and the breakthrough of aid decision making on theory, method, technology and function.2) set up one and had versatility, opening, extensibility, reusable, based on the Predicting and Policy-Making integral intelligent simulating experimental system model of multi-Agent technology.3) exploitation one cover powerful, easy-to-use, integrated, visual, cheaply, calculate and find the solution ability high-intelligentization decision simulating experimental software systems surpassingly.4) first isomery Agent and group collaboration thereof are introduced DSS, increased substantially level of intelligence; With the integrated processing of Predicting and Policy-Making technology, increased substantially the science of decision-making first; First the cloud computing business model is applied to aid decision making, increased substantially the ageing of decision-making.5) make up the architecture of intelligent decision simulating experimental system, not only changed the operating mechanism of existing analogue system, more changed the building method and the operating mechanism of existing intelligent decision support system.6) design and development of intelligent decision simulating experimental system has been brought up to the strategy management aspect with traditional for Operations Management designed system emulation technology; And will be all the time independent exist and China's management experimental study of slowly development has been brought internet and intelligent body epoch into, also will cause that systems engineering bound pair management decision problem crossing domain studies widely simultaneously.
Description of drawings
Below in conjunction with accompanying drawing and embodiment the present invention is described further.
Fig. 1 is intelligent decision simulating experimental system (IDSES) structural representation that the present invention designs.
Fig. 2 is four layers of Agent architectural schematic of IDSES that the present invention designs.
Fig. 3 is the IDSES planning subsystem work process flow diagram that the present invention designs.
Fig. 4 is the IDSES evaluation subsystem workflow diagram that the present invention designs.
Fig. 5 is the IDSES decision-making subsystem workflow diagram that the present invention designs.
Fig. 6 is the IDSES game subsystem work process flow diagram that the present invention designs.
Fig. 7 is function layer MAS man-machine interface Agent structure and the functional schematic that the present invention designs
Fig. 8 is that the function layer MAS that the present invention designs explains Agent structure and functional schematic
Fig. 9 is function layer MAS manual intervention Agent structure and the functional schematic that the present invention designs
Figure 10 is function layer MAS constraint Agent structure and the functional schematic that the present invention designs
Figure 11 is function layer MAS reasoning Agent structure and the functional schematic that the present invention designs
Figure 12 is work layer MAS planning Agent structure and the functional schematic that the present invention designs
Figure 13 is work layer MAS assessment Agent structure and the functional schematic that the present invention designs
Figure 14 is work layer MAS decision analysis Agent structure and the functional schematic that the present invention designs
Figure 15 is work layer MAS game Agent structure and the functional schematic that the present invention designs
Figure 16 is work layer MAS task Agent structure and the functional schematic that the present invention designs
Figure 17 is work layer MAS target Agent structure and the functional schematic that the present invention designs
Figure 18 is work layer MAS strategy Agent structure and the functional schematic that the present invention designs
Figure 19 is work layer MAS project Agent structure and the functional schematic that the present invention designs
Figure 20 is work layer MAS operation Agent structure and the functional schematic that the present invention designs
Figure 21 is work layer MAS Information Agent structure and the functional schematic that the present invention designs
Figure 22 is work layer MAS natural language processing Agent structure and the functional schematic that the present invention designs
Figure 23 is work layer MAS emulation Agent structure and the functional schematic that the present invention designs
Figure 24 is work layer MAS graphic modeling Agent structure and the functional schematic that the present invention designs
Figure 25 is communication layers MAS communication Agent structure and the functional schematic that the present invention designs
Figure 26 is resource layer MAS material database Management Agent structure and the functional schematic that the present invention designs
Figure 27 is resource layer MAS KBM Agent structure and the functional schematic that the present invention designs
Figure 28 is resource layer MAS logical base Management Agent structure and the functional schematic that the present invention designs
Figure 29 is resource layer MAS example/countermeasure library management Agent structure and functional schematic that the present invention designs
Figure 30 is that the resource layer MAS teledata that the present invention designs is excavated Agent structure and functional schematic
Embodiment
As shown in Figure 1, be designed system structural drawing of the present invention.IDSES is made up of four sub-systems, and each subsystem is realized predetermined function through four layers of MAS.System's desired data is stored in four databases, and the logging in system by user interface is selected to get into different subsystems as required and operated, and data are shared between each subsystem.When data deficiencies, system initiates teledata and excavates process, obtains desired data from Internet; When system's generation large-scale parallel computation requirement, can utilize the cloud computing resource to improve through Internet and find the solution efficient.
Four layers of Agent architectural schematic as shown in Figure 2, as to design for the present invention.A plurality of isomery Agent are divided into four levels according to function needs and mutual needs.The function layer has comprised 5 kinds of Agent in four functional subsystems, is and the immediate upper strata of user performance mechanism, is used to set up human-computer interaction and carries out and the auxiliary correction work of conclusion, has directly showed the function of system; Work layer has comprised 13 kinds of Agent, is the middle core topworks of system, is transparent for the user, and the instruction of their upper layer issues is carried out a series of solution procedurees through communications intermediary from resource layer acquisition data and then final structure fed back to the function layer; Communication layers has only communication Agent, is the middle level communication backbone of system, mainly is used for pass-along message and data and carry out Agent registration, guarantees communications and data transmits in the Agent system order and validity; Resource layer comprises 6 kinds of Agent, is the underlying resource management organization of system, is used for the analysis and the identification of Data Acquisition, renewal, submission and mass data, guarantees the accuracy of the frequent invoked procedure of data.
Planning subsystem work process flow diagram as shown in Figure 3, as to design for the present invention.Its concrete steps are following:
Step 1, the user need formulate a scheme, through man-machine interface login planning subsystem, selects machine planning or artificial planning, and each Agent of planning subsystem is to communication Agent registration; If selected machine planning, then in the input of inputting interface choice structure text or non-structured text input problem and parameter; If selected artificial planning, the figure material that then carries out textual description problem and parameter or utilize system to provide at inputting interface carries out graphical modeling and describes logic of questions at the graphical modeling platform;
Step 2, the information of input carries out information filtering through natural language processing Agent and crucial words and phrases extract, and converts the normal data of system identification to;
Step 3, if selection is machine planning, normal data will be directly passed to task Agent, and task Agent at first inquires about corresponding separating through communication Agent to case library, and in the hope of most effective, communication Agent carries out message queueing and forwarding;
Step 4, case library Management Agent are inquired about corresponding example to case library after receiving the message that communication Agent transmits, if example does not exist and then do not return null value and give communication Agent, communication Agent transmits this message and gives task Agent; If existing, example then it is submitted to natural language processing Agent; Natural language processing Agent receives that example is handled and gives communication Agent with data;
Step 5; In the machine planning route, if example Agent returns null value, then task Agent carries out the task decomposition according to the character of customer problem and parameter and the ability of projects Agent; And the subtask of collecting projects Agent separates the formation problem and separates, and submits to planning Agent then; If what select is artificial planning, normal data will pass to planning Agent;
Step 6; In the machine planning route; Project Agent accepts the subtask that task Agent distributes, according to task difficulty and workload decision and other project Agent cooperation or further the operation exercise question is resolved in the subtask and give operation Agent and collect exercise question that operation Agent submits to and separate the formation subtask and separate and submit to the task Agent of distributing to own subtask;
Step 7; In the machine planning route; Operation Agent accepts the operation exercise question and sends demand data message according to the data of operation exercise question needs to Information Agent, obtains separating according to derivation algorithm computational tasks exercise question after the data, and will separate and submit to the project Agent that distributes to own operation exercise question;
Step 8, in the machine planning route, Information Agent is sent the desired data parameter to communication Agent after receiving the message of demand data, and communication Agent give KBM Agent and logical base Management Agent with this forwards;
Step 9; In the machine planning route; KBM Agent and logical base Management Agent are received after the message of demand data to knowledge base and logical base inquiry desired data; Submit to Information Agent if desired data exists then to transmit through communication Agent, Information Agent is submitted to the data that obtain the operation Agent that asks for instructions to oneself sending; If desired data does not exist KBM Agent then and/or logical base Management Agent to excavate the message that Agent sends the data shortage to teledata;
Step 10; In the machine planning route; Initiation data mining process was carried out Knowledge Discovery through LAN or internet after teledata excavation Agent received the short message of data; If desired data can find then it is submitted to communication Agent and deposits corresponding database in, if desired data can not find then message is submitted to manual intervention Agent;
Step 11; In the machine planning route; Manual intervention Agent receives that data mining is to start editing machine and/or compiler program after the empty message; Carry out replenishing of artificial knowledge, logic by user or professional, and the data of finishing are submitted to Information Agent through communication Agent, deposit corresponding database simultaneously in;
Step 12, the method in existing planning principles of planning Agent basis and planning territory are carried out standard to the artificial planning scheme and are formed the standard planning scheme or the machine planning problem is separated formation standard planning scheme, and this scheme is submitted to constraint Agent;
Step 13, constraint Agent be according to carrying out constraint Analysis with bounding algorithm and user's constrained conditions set to programme, or scheme meets and then submit to emulation Agent, if scheme does not meet then this message and scheme are submitted to reasoning Agent;
Step 14, reasoning Agent carries out rational analysis to scheme and suggests improvements submitting to planning Agent according to corresponding reasoning algorithm and expertise, logic;
Step 15, planning Agent receives after the corresponding suggestion scheme revised and retrains comparison once more that qualified scheme will be submitted to emulation Agent;
Step 16; Emulation Agent receives that programme carries out the planning implementation process simulation and calculate the result according to simulation algorithm; If user's decision does not need graphic modeling; Then scheme and simulation result are submitted in the lump and are explained Agent, if the user determines to carry out graphic modeling, then scheme is submitted to graphic modeling Agent;
Step 17; Graphic modeling Agent sends the material demand through communication Agent to the material database Management Agent after receiving the scheme that needs simulation; The material database Management Agent is inquired about required material to material database after receiving requirement message; Submit to graphic modeling Agent if material exists then through communication Agent, do not excavate the message that Agent sends the data shortage if material does not exist then to teledata, process is with knowledge and logical data shortage request process;
Step 18, graphic modeling Agent obtains behind the required material carrying out two dimension or three-dimensional picture dynamic similation according to programme and corresponding action algorithm to be showed simultaneously simulation process to be passed to and explains Agent;
Step 19; Explain after Agent receives simulation result and/or graphic modeling process and final plan is carried out literal interpretation according to interpretation algorithms and expertise; Programme is explained that with following exporting to man-machine interface Agent deposits case library simultaneously in, feeds back to the user then.
As shown in Figure 4, be the evaluation subsystem workflow diagram that the present invention designs, its concrete steps are following:
Step 1, the user is through man-machine interface login evaluation subsystem, and each Agent of evaluation subsystem is to communication Agent registration; The user carries out graphical modeling and describes the prediction scheme logic at the graphical modeling platform in inputting interface choice structure text input prediction scheme or non-structured text input prediction scheme or the figure material that utilizes system to provide;
Step 2, the prediction scheme of input carries out information filtering through natural language processing Agent and crucial words and phrases extract, and converts the normal data of system identification to and submits to emulation Agent;
Step 3; Emulation Agent receives that prediction scheme carries out the emulation of prediction scheme implementation process and calculate the result according to simulation algorithm; If user's decision does not need graphic modeling; Then prediction scheme and simulation result are submitted to assessment Agent in the lump, if the user determines to carry out graphic modeling, then scheme is submitted to graphic modeling Agent;
Step 4; Graphic modeling Agent sends the material demand through communication Agent to the material database Management Agent after receiving the prediction scheme that needs simulation; The material database Management Agent is inquired about required material to material database after receiving requirement message; Submit to graphic modeling Agent if material exists then through communication Agent, do not excavate the message that Agent sends the data shortage if material does not exist then to teledata, process is with the knowledge and the logical data shortage process of planning subsystem;
Step 5, graphic modeling Agent obtains behind the required material composition, logic and corresponding actions algorithm according to prediction scheme to carry out two dimension or three-dimensional picture dynamic similation and shows simultaneously simulation process is passed to assessment Agent;
Step 6 after assessment Agent receives prediction scheme simulation process and result, is carried out the prediction scheme evaluation according to expertise and user's constraint condition, and is combined assessment algorithm, expertise and ownership goal to calculate more excellent separating, and forms comments and gives explanation Agent; The required knowledge of assessment Agent, logic are directly asked for to underlying resource management MAS through communication Agent, and the bottom data inquiry is identical to step 15 with planning subsystem step 12 with the submission process;
Step 7 is explained after Agent receives the prediction scheme comments and according to interpretation algorithms and expertise prediction scheme and comments is carried out literal interpretation, with comments with follow explanation to export to man-machine interface Agent and deposit case library in, feeds back to the user then.
As shown in Figure 5, be the decision-making subsystem workflow diagram that the present invention designs, its concrete steps are following:
Step 1, the user is through man-machine interface login decision-making subsystem, and each Agent of decision-making subsystem is to communication Agent registration; The user imports 2 above prediction schemes successively or utilizes figure material that system provides to carry out graphical modeling and describe the logic of each prediction scheme at the graphical modeling platform at inputting interface choice structure text input or non-structured text;
Step 2, each prediction scheme of input carry out information filtering through natural language processing Agent successively and crucial words and phrases extract, and converts the normal data of system identification to and submits to emulation Agent;
Step 3; Emulation Agent receives that each prediction scheme carries out implementation process emulation successively and calculate the result each prediction scheme according to simulation algorithm; If user's decision does not need graphic modeling; Then each prediction scheme and simulation result thereof are submitted to decision analysis Agent in the lump, if the user determines to carry out graphic modeling, then each prediction scheme is submitted to graphic modeling Agent successively;
Step 4; Graphic modeling Agent sends the material demand through communication Agent to the material database Management Agent after receiving each prediction scheme that needs simulation; The material database Management Agent is inquired about required material to material database after receiving requirement message; Submit to graphic modeling Agent if material exists then through communication Agent, do not excavate the message that Agent sends the data shortage if material does not exist then to teledata, process is with the knowledge and the logical data shortage process of planning subsystem;
Step 5, graphic modeling Agent obtains behind the required material composition, logic and corresponding action algorithm according to each prediction scheme to carry out two dimension or three-dimensional picture dynamic similation and shows and simultaneously simulation process is passed to decision analysis Agent successively;
Step 6; After decision analysis Agent receives each prediction scheme simulation process and result; Carry out each prediction scheme comparative analysis according to expertise and user's constraint condition, combine decision making algorithm, expertise and ownership goal to calculate optimum solution simultaneously, form the decision analysis suggestion and submit to explanation Agent; The required knowledge of decision analysis Agent, logic are directly asked for to underlying resource management MAS through communication Agent, and the bottom data inquiry is identical to step 15 with planning subsystem step 12 with the submission process;
Step 7; Explain that Agent receives according to interpretation algorithms and expertise each prediction scheme and decision analysis suggestion being carried out literal interpretation after the decision analysis suggestion of each prediction scheme; The decision analysis suggestion is explained that with following exporting to man-machine interface Agent deposits case library simultaneously in, feeds back to the user then.
As shown in Figure 6, be the game subsystem work process flow diagram that the present invention designs, its concrete steps are following:
Step 1, the user need check the feasibility and the competitive power of decision-making, through man-machine interface login game subsystem, selects man-machine game or online game, and each Agent of game subsystem is to communication Agent registration; No matter which kind of game adds parameter at inputting interface with semi-structured text input problem, with structuring or non-structured text input policing;
Step 2, the information of input carries out information filtering through natural language processing Agent and crucial words and phrases extract, and converts the normal data of system identification to;
Step 3; Normal data will be directly passed to game Agent; If what the user selected is man-machine game, then game Agent generates countermeasure with customer problem, parameter and strategy and finds the solution information, at first separates to the countermeasure library inquiry is corresponding through communication Agent; In the hope of most effective, communication Agent carries out message queueing and forwarding;
Step 4, countermeasure library management Agent receives after the message that communication Agent transmits to countermeasure library inquiry Corresponding Countermeasures, if countermeasure does not exist and then do not return null value and give communication Agent, communication Agent transmits this message and gives game Agent; If existing, countermeasure then it is submitted to natural language processing Agent;
Step 5, natural language processing Agent receives that countermeasure carries out information processing, and the data after the processing are given communication Agent, and communication Agent give game Agent with this data forwarding;
Step 6; In man-machine game route; If countermeasure library management Agent returns null value, then game Agent finds the solution information with countermeasure and gives tactful Agent, and tactful Agent carries out the task decomposition according to the character of customer problem, parameter and the ability of tactful intensity and projects Agent; The subtask of giving the project Agent that chooses and collecting projects Agent is separated and is formed countermeasure and separate, and submits to game Agent then; In online game route; Game Agent finds the solution information with countermeasure and gives target Agent; Target Agent seeks predetermined game object through the Agent that communicates by letter to LAN or wide area network with tactful intensity and algorithms for searching objects according to the character of customer problem, parameter, and its strategy is submitted to game Agent;
Step 7; In man-machine game route; Project Agent accepts the subtask that tactful Agent distributes, according to task difficulty and workload decision and other project Agent cooperation or further the operation exercise question is resolved in the subtask and give operation Agent and collect exercise question that operation Agent submits to and separate the formation subtask and separate and submit to the tactful Agent that distributes to own subtask;
Step 8; In man-machine game route; Operation Agent accepts the operation exercise question and sends demand data message according to the data of operation exercise question needs to Information Agent, obtains separating according to derivation algorithm computational tasks exercise question after the data, and will separate and submit to the project Agent that distributes to own operation exercise question;
Step 9; In man-machine game route; Information Agent is sent the desired data parameter to communication Agent after receiving the message of demand data, ask for to underlying resource management MAS through communication Agent, bottom data inquiry and submission process with plan that subsystem step 12 is identical to step 15;
Step 10; Game Agent basis is problem+parameter environment, ruling model and game playing algorithm at that time; Be aided with the expertise reasoning in case of necessity the expansion and the implementation process of subscriber policy and machine policy or adversary's strategy are carried out critical data point collection judgement and ruling, tactful implementation process and ruling process are submitted to emulation Agent;
Step 11; Emulation Agent carries out tactful implementation process emulation and calculates the ruling result according to simulation algorithm after receiving both sides' strategy implementation process and ruling process; If user's decision does not need graphic modeling; Then machine policy or adversary strategy and simulation result are submitted in the lump and are explained Agent, if the user determines to carry out graphic modeling, then submit to graphic modeling Agent;
Step 12; Graphic modeling Agent sends the material demand through communication Agent to the material database Management Agent after receiving the object that needs simulation; The material database Management Agent is inquired about required material to material database after receiving requirement message; Submit to graphic modeling Agent if material exists then through communication Agent, do not excavate the message that Agent sends the data shortage if material does not exist then to teledata, process is with the knowledge and the logical data shortage process of planning subsystem;
Step 13, graphic modeling Agent obtains behind the required material carrying out two dimension or three-dimensional picture dynamic similation according to tactful game and ruling process and corresponding actions algorithm to be showed simultaneously simulation process to be passed to and explains Agent;
Step 14; Explain after Agent receives simulation result and/or graphic modeling process and final payoff is carried out literal interpretation according to interpretation algorithms and expertise; With machine policy or adversary strategy, payoff with follow and explain and export to man-machine interface Agent (depositing machine policy in the countermeasure storehouse simultaneously), feed back to the user then.
As shown in Figure 7, the function layer MAS man-machine interface Agent structure and the functional schematic that design for the present invention.Deposit other Agent information in the Agent state table, can use uncertain logical description.The Agent dictionary is deposited user-defined standardization representation of concept and explanation, and man-machine interface Agent is the input and output of update information in view of the above.Input/output control module perceives task, carries out input and output control according to dictionary and I/O algorithm, searches the Agent state table in case of necessity, reports an error or requires to revise the I/O algorithm to the user.Communication module be responsible for man-machine interface A gent registration and with the communicating by letter of other Agent, the communication module function of other Agent is identical therewith, repeats no more.
As shown in Figure 8, the function layer MAS that designs for the present invention explains Agent structure and functional schematic.Deposit other Agent information in the Agent state table, can use uncertain logical description.The Agent dictionary is deposited user-defined standardization representation of concept and explanation, explains that Agent can revise the explanation conclusion in view of the above.Explain that maker perceives task, make an explanation, search the Agent state table in case of necessity, grasp global process, form the correct interpretation conclusion according to dictionary and logical base, knowledge base.The conclusion control module is according to explanation task and interpretation algorithms execution control function.If in interpretation process, find data shortage or concept obfuscation, the conclusion control module will be excavated the Agent request of sending or require the user to upgrade the Agent dictionary to teledata.Interpretation process possibly need user intervention.
As shown in Figure 9, the function layer MAS manual intervention Agent structure and the functional schematic that design for the present invention.The routine call control module perceives task, carries out application call according to calling algorithm.
Shown in figure 10, the function layer MAS constraint Agent structure and the functional schematic that design for the present invention.Deposit constraint condition information in the constraint condition table, with confirming logical description.The Agent dictionary is deposited user-defined standardization representation of concept and explanation, and constraint Agent can be revised constraint condition in view of the above.Bound control module perceives task, searches the Agent state table, carries out constraint test according to dictionary and bounding algorithm, communication module be responsible for constraint Agent registration and with the communicating by letter of other Agent.
Shown in figure 11, the function layer MAS reasoning Agent structure and the functional schematic that design for the present invention.Rule is left decision-making and planning experiment, available definite logical description concentratedly.The Agent dictionary is deposited user-defined standardization representation of concept and explanation, and reasoning Agent can revise reasoning process and conclusion in view of the above.The problem deduce machine perceives task, carries out reasoning according to dictionary and logical base, knowledge base, searches the planning collection in case of necessity, forms correct reasoning suggestion.The rule control module is problem and reasoning algorithm execution control function by inference, like coupling, conflict resolution etc.If in reasoning process, find data shortage or logical miss, regular control module will be excavated the Agent request of sending or require the user to upgrade the Agent dictionary to teledata.Reasoning process possibly need user intervention.
Shown in figure 12, the work layer MAS that designs for the present invention plans Agent structure and functional schematic.Deposit other Agent information in the Agent state table,, can use uncertain logical description like problem solving scope, error rate etc.The Agent dictionary is deposited user-defined standardization representation of concept and explanation, and planning Agent can revise mission planning in view of the above.The mission planning device perceives task, carries out mission planning according to dictionary and case library, logical base, knowledge base, searches the Agent state table in case of necessity, selects suitable task or project Agent to bear corresponding subtask, forms complete planning.Decision-making and intelligent control module are according to mission planning and planning algorithm execution control function.If in planning process, find the data shortage, decision-making will be excavated Agent to teledata with intelligent control module and sent request.Possibly need user intervention in the planning process.
Shown in figure 13, the work layer MAS that designs for the present invention assesses Agent structure and functional schematic.Deposit other Agent information in the Agent state table, can use uncertain logical description.The Agent dictionary is deposited user-defined standardization representation of concept and explanation, and assessment Agent can revise comments in view of the above.The evaluate alternatives module perceives task, carries out scheme evaluation according to dictionary and logical base, knowledge base, user's constraint condition, searches the Agent state table in case of necessity, grasps full information, forms correctly and estimates.The assessment control module is according to evaluate alternatives and assessment algorithm execution control function.If in evaluation process, find the data shortage, the assessment control module will be excavated Agent to teledata and send request.Possibly need user intervention in the evaluation process.
Shown in figure 14, the work layer MAS decision analysis Agent structure and the functional schematic that design for the present invention.Deposit other Agent information in the Agent state table, can use uncertain logical description.The Agent dictionary is deposited user-defined standardization representation of concept and explanation, and decision analysis Agent can revise the decision analysis conclusion in view of the above.The scheme comparison module perceives task, carries out scheme relatively according to dictionary and logical base, knowledge base, user's constraint condition, searches the Agent state table in case of necessity, grasps full information, forms the correct analysis conclusion.The decision analysis control module compares and the decision making algorithm execution control function according to scheme.If in decision analysis process, find the data shortage, the decision analysis control module will be excavated Agent to teledata and send request.Possibly need user intervention in the decision analysis process.
Shown in figure 15, the work layer MAS game Agent structure and the functional schematic that design for the present invention.Deposit other Agent information in the Agent state table,, can use uncertain logical description like problem solving scope, error rate etc.The Agent dictionary is deposited user-defined standardization representation of concept and explanation, and game Agent can revise ruling process and conclusion in view of the above.Strategy generator perceives task, carries out the strategy generation according to dictionary and countermeasure storehouse, logical base, knowledge base, searches the Agent state table in case of necessity, selects suitable strategy or project Agent to bear corresponding subtask, forms complete strategy.Ruling and intelligent control module generate and the game playing algorithm execution control function according to strategy.If in the game process, find data shortage and logical miss, ruling and intelligent control module will excavate the Agent request of sending and/or require the user to upgrade dictionary to teledata.Possibly need user intervention in the game process.
Shown in figure 16, the work layer MAS task Agent structure and the functional schematic that design for the present invention.Deposit other Agent information in the Agent state table,, can use uncertain logical description like problem solving scope, error rate etc.The Agent dictionary is deposited user-defined standardization representation of concept and explanation, and task Agent can revise the subtask distribution in view of the above and the subtask is separated synthetic.The project planning device perceives task, under the control of planning Agent, carries out the subtask according to dictionary and logical base, knowledge base and decomposes and conciliate syntheticly, searches the Agent state table, selects suitable project Agent to bear corresponding subtask, forms complete planning.Decision-making and intelligent control module are according to project planning and planning algorithm execution control function.If in planning process, find the data shortage, decision-making will be excavated Agent to teledata with intelligent control module and sent request.
Shown in figure 17, the work layer MAS target Agent structure and the functional schematic that design for the present invention.Deposit LAN or wide area network destination path information in the routing table.The Agent dictionary is deposited user-defined standardization representation of concept and explanation, and target Agent can revise search condition and path in view of the above.The target search control module perceives task, searches routing iinformation, carries out target search according to dictionary and searching algorithm.
Shown in figure 18, the work layer MAS strategy Agent structure and the functional schematic that design for the present invention.Deposit other Agent information in the Agent state table,, can use uncertain logical description like problem solving scope, error rate etc.The Agent dictionary is deposited user-defined standardization representation of concept and explanation, and tactful Agent can revise the subtask distribution in view of the above and countermeasure is synthesized.The project planning device perceives task, carries out under the control of game Agent according to dictionary and logical base, knowledge base that decompose the subtask and countermeasure is synthetic, searches the Agent state table, selects suitable project Agent to bear corresponding subtask, forms complete strategy.The game control module is according to project planning and planning algorithm execution control function.If in tactful generative process, find the data shortage, the game control module will be excavated Agent to teledata and send request.
Shown in figure 19, the work layer MAS project Agent structure and the functional schematic that design for the present invention.Deposit other Agent information in the Agent state table, can use uncertain logical description.The Agent dictionary is deposited user-defined standardization representation of concept and explanation, and project Agent can carry out in view of the above that subproblem is found the solution or carry out subproblem operation exercise question and distribute and separate synthetic.The work planning device perceives task; According to dictionary and logical base, knowledge base under the control of task Agent, carry out the operation exercise question decompose conciliate synthetic; Search the Agent state table in case of necessity, select suitable operation Agent, form complete subproblem and separate to bear corresponding operation exercise question.Decision-making and intelligent control module are according to work planning and planning algorithm execution control function.If in solution procedure, find the data shortage, decision-making will be excavated Agent to teledata with intelligent control module and sent request.
Shown in figure 20, the work layer MAS operation Agent structure and the functional schematic that design for the present invention.Deposit other Agent information in the Agent state table, can use uncertain logical description.The Agent dictionary is deposited user-defined standardization representation of concept and explanation, and operation Agent can carry out in view of the above and revise subproblem and find the solution and separate synthetic.The subproblem solver perceives task; According to dictionary and logical base, knowledge base under the control of project Agent, carry out subproblem find the solution conciliate synthetic; Search the Agent state table in case of necessity, select suitable operation Agent to accomplish problem solving, form complete subproblem and separate with cooperation.Find the solution with intelligent control module and find the solution and the derivation algorithm execution control function according to subproblem.If in solution procedure, find the data shortage, find the solution and to excavate Agent to teledata with intelligent control module and send request.
Shown in figure 21, the work layer MAS Information Agent structure and the functional schematic that design for the present invention.Deposit other Agent information in the Agent state table, can use uncertain logical description.The Agent dictionary is deposited user-defined standardization representation of concept and explanation, and Information Agent is update information data layout and expression way in view of the above.The information acquisition control module perceives task, carries out information inquiry and collection according to dictionary and collection algorithm, searches the Agent state table in case of necessity, with out of Memory Agent cooperation, forms the complete information data.
Shown in figure 22, the work layer MAS natural language processing Agent structure and the functional schematic that design for the present invention.Deposit the proprietary lexical information of industry in the proprietary vocabulary.User-defined standardization syntactic representation deposited by the Agent dictionary and ambiguity is explained, natural language processing Agent is processing procedure such as update information retrieval and keyword abstraction in view of the above.Language processing module perceives task, carries out the natural language text processing according to dictionary and corpus, logical base, knowledge base, extracts key message and forms standardized data, searches proprietary vocabulary in case of necessity, selects suitable vocabulary, forms the correct handling result.Information Recognition and intelligent control module are according to Language Processing and recognizer execution control function, like error correction, ambiguity analysis, the not enough prompting of information etc.If in the Language Processing process, find data shortage or identification difficulty, information Recognition and intelligent control module will to teledata excavate Agent and/manual intervention Agent sends request.The natural language processing process possibly need user intervention.
Shown in figure 23, the work layer MAS emulation Agent structure and the functional schematic that design for the present invention.Deposit function expression and attribute thereof in the function table.The Agent dictionary is deposited user-defined standardization representation of concept and explanation, and emulation Agent can revise simulation process in view of the above.The Simulation Control module perceives task, carries out the process simulation of scheme and submits simulation result to according to dictionary and simulation algorithm, searches function table in case of necessity, selects suitable construction of function realistic model, forms correct emulation conclusion.
Shown in figure 24, the work layer MAS graphic modeling Agent structure and the functional schematic that design for the present invention.Deposit graphical modeling function and attribute thereof in the function table, graphic modeling Agent can set up graphic scheme framework and mutual relationship in view of the above.The Agent dictionary is deposited user-defined standardization representation of concept and explanation, and graphic modeling Agent can revise role and movement locus thereof in view of the above.The graphic modeling control module perceives task, carries out the dynamic graphical simulation shows of scheme according to dictionary and action algorithm.
Shown in figure 25, the communication layers MAS communication Agent structure and the functional schematic that design for the present invention.Deposit the log-on message of all Agent in the registration table, can use definite logical description, the information that communication Agent could discern and search target Agent in view of the above.Routing table is deposited forwarding information queueing condition and path seizure condition, and communication Agent can find the suitable path information of carrying out to transmit in view of the above rapidly.The communication server perceives information, transmits according to registration table, routing table and the communication of algorithms information of carrying out, and message transmission module is responsible for packet and is sent.
Shown in figure 26, the resource layer MAS material database Management Agent structure and the functional schematic that design for the present invention.Deposit other Agent information in the Agent state table, can use uncertain logical description.The Agent dictionary is deposited user-defined standardization representation of concept and explanation, and the material database Management Agent can be revised the mode classification and the memory management policy of material in view of the above.Material is classified and is called control module and perceives task; According to dictionary and classification with call that algorithm carries out material call, store classification; Search the Agent state table when data shortage or inefficacy, log-on data excavation, manual intervention and redundant clean up process, this process possibly need and user interactions.
Shown in figure 27, the resource layer MAS KBM Agent structure and the functional schematic that design for the present invention.Deposit other Agent information in the Agent state table, can use uncertain logical description.The Agent dictionary is deposited user-defined standardization representation of concept and explanation, and KBM Agent can carry out knowledge-based inference in view of the above and revise knowledge classification mode, storage policy, inference logic.Knowledge classification with call control module and perceive task, according to classification with call algorithm data-base and carry out the calling of knowledge, classification and storage, and relatively update module is mutual with knowledge, to guarantee the freshness of knowledge.Knowledge comparison update module is carried out the comparing function based on time and user interpretation according to dictionary and comparison algorithm to the data that deposit knowledge base in; When knowledge shortage and inefficacy, search the Agent state table; Log-on data excavation, manual intervention and redundant clean up process, this process possibly need and user interactions.Relatively the update module circulation is mutual for knowledge-based inference module and Agent dictionary, knowledge, guarantees the self of knowledge and dictionary, and deposits knowledge-based inference process and conclusion in knowledge base as new knowledge.
Shown in figure 28, the resource layer MAS logical base Management Agent structure and the functional schematic that design for the present invention.Deposit other Agent information in the Agent state table, can use uncertain logical description.The Agent dictionary is deposited user-defined standardization representation of concept and explanation, and the logical base Management Agent can be carried out reasoning and correction logic mode classification, storage policy, the reasoning principle of logic-based in view of the above.Logical division with call control module and perceive task, according to classification with call algorithm data-base and carry out the calling of logic, classification and storage, and relatively update module is mutual with logic, to guarantee the advance and the applicability of logic.Logic comparison update module is carried out the comparing function based on time and user interpretation according to dictionary and comparison algorithm to the data that deposit logical base in; When logic shortage and inefficacy, search the Agent state table; Log-on data excavation, manual intervention and redundant clean up process, this process possibly need and user interactions.Relatively the update module circulation is mutual for the reasoning module of logic-based and Agent dictionary, logic, guarantees the self of logic and dictionary, and deposits the reasoning process and the conclusion of logic-based in logical base as new logic.
Resource layer MAS example/countermeasure library management Agent structure and functional schematic shown in figure 29, as to design for the present invention.Deposit other Agent information in the Agent state table, can use uncertain logical description.The Agent dictionary is deposited user-defined standardization representation of concept and explanation, and example/countermeasure library management Agent can revise example/classification of game mode, storage policy, comparison condition in view of the above.Example/classification of game with call control module and perceive task, according to classification with call algorithm data-base and carry out the calling of example/countermeasure, classification and storage, and relatively update module is mutual with example/countermeasure, to guarantee the freshness of example/countermeasure.Example/countermeasure comparison update module is carried out the comparing function based on time and user interpretation according to dictionary and comparison algorithm to the data that deposit example/countermeasure storehouse in; When example/countermeasure shortage and inefficacy, search the Agent state table; Log-on data excavation, manual intervention and redundant clean up process, this process possibly need and user interactions.
Shown in figure 30, for the resource layer MAS teledata that the present invention designs is excavated Agent structure and functional schematic.Deposit the resource information of cloud computing resource and network node data in the Internet resources retrieval,, teledata is excavated Agent can rapid in view of the above localizing objects scope.The Agent dictionary is deposited user-defined standardization representation of concept and explanation, and teledata is excavated Agent can revise data mining process in view of the above.The data mining control module perceives task, from the KDD component base, calls suitable member according to data type, inserts Internet through communication Agent and carries out the teledata excavation.After obtaining data, the evaluation of Agent log-on data with test module and the data and the Relational database that are obtained checked according to dictionary, find out and clear up redundancy and conflict, this process possibly need and user interactions.Data of different types is needed different KDD algorithms, adopt Knowledge Discovery component base mechanism, various knowledge discovery algorithms are made member leave in the component base, call from component base according to data type by the data mining control module.
Claims (4)
1. intelligent decision simulating experimental system based on multi-Agent technology, english abbreviation is IDSES, it be one based on internet, integrated Predicting and Policy-Making integral intelligent analogue system; It is characterized in that:
(1) said intelligent decision simulating experimental system comprises planning subsystem, evaluation subsystem, decision-making subsystem, four functional subsystems of game subsystem;
(2) said each functional subsystem has four layers of Agent system structure, and english abbreviation is MAS, promptly functional layer MAS, work layer MAS, communication layers MAS, resource layer MAS;
(3) said planning subsystem, its function are to carry out intelligence machine planning or artificial planning and propose decision scheme;
Function layer MAS comprises:
Man-machine interface Agent is used for obtaining information, to the user conclusion and explanation facility is provided, and the request user imports necessary information;
Explain Agent, be used for experimental result is carried out necessary, rational explanation to operational process, feed back to man-machine interface Agent to the result and export to the user;
Manual intervention Agent is used for when inadequate resource, starting corresponding program, is handled by user intervention;
Constraint Agent is used to solve the constraint satisfaction of decision experiments process, mainly solves common-sense and restricted problem agreement;
Reasoning Agent relies on expertise and strategy that the machine planning scheme is carried out true inference, and Agent suggests improvements to planning;
Work layer MAS comprises:
Planning Agent is used for accepting user's planning experiment request, sets up planning constraint, plan model and planning territory, and the program results of integrated each task Agent forms the general planning case;
Task Agent under planning Agent control, carries out PROBLEM DECOMPOSITION according to projects Agent ability and state, gives corresponding project Agent Task Distribution, and the sub-case of planning of assembled item Agent feeds back to planning Agent simultaneously;
Project Agent is a main body of accomplishing experimental duties, bears concrete experimental duties, can derive down one deck operation Agent;
Operation Agent is a base unit of accomplishing experimental duties, according to resulting subproblem and parameter, utilizes relevant information formation base experimental result and feeds back to project Agent;
Information Agent be responsible for to be collected required various information of upper level Agent and feedback, needs a plurality of Information Agent to accomplish various tasks and to hold consultation and cooperate;
Natural language processing Agent is used for process user at experimental system interface input and extensive real text system's output, carries out information retrieval filtration and information extraction, and the succinct authority data that converts system identification to is handled to treat core Agent;
Emulation Agent is used for the standardization decision scheme is carried out mathematical simulation, calculating and reasoning from logic emulation and exports the result;
Graphic modeling Agent is used for simulation process is simulated with suitable graph visualization, has spatial parameter and behavior calculation of parameter function simultaneously;
Communication layers MAS comprises:
Communication Agent in order to realize data queue, provides the registration of Agent, message based communication, and the supplier provides coupling for message, the interpretation process function;
Resource layer MAS comprises:
The material database Management Agent is responsible for management maintenance, the material combination of material database, the material operation;
KBM Agent is responsible for MAINTENANCE OF KNOWLEDGE BASE, knowledge is relatively upgraded and knowledge-based inference work;
The logical base Management Agent, maintenance, the logic of being responsible for logical base called, logic is mated and the reasoning work of logic-based;
Example/countermeasure library management Agent is responsible for the example/retrieval in countermeasure storehouse, coupling, modification, multiplexing corresponding function based on case-based reasoning;
Teledata is excavated Agent, is responsible for carrying out when local data lacks that teledata is obtained, data are tested and data collision and redundancy are cleared up;
(4) said evaluation subsystem, its function are that the decision-making prediction scheme is carried out intelligent evaluation and suggested improvements;
Function layer MAS comprises: man-machine interface Agent, explanation Agent, manual intervention Agent;
Work layer MAS comprises: natural language processing Agent, emulation Agent, graphic modeling Agent, assessment Agent; Wherein assessing Agent is the assessment experiment request that is used for accepting the user, sets up evaluation index system and test set, and the target that the decision scheme user is concerned about compares evaluation and reasoning from logic evaluation, forms evaluation of programme;
Communication layers MAS comprises: communication Agent;
Resource layer MAS comprises: material database Management Agent, KBM Agent, logical base Management Agent, example/countermeasure library management Agent, teledata are excavated Agent;
(5) said decision-making subsystem, its function are that a plurality of decision-making prediction schemes are carried out intellectual analysis and selected satisfactory solution;
Function layer MAS comprises: man-machine interface Agent, explanation Agent, manual intervention Agent;
Work layer MAS comprises: natural language processing Agent, emulation Agent, graphic modeling Agent, decision analysis Agent; Wherein decision analysis Agent is the decision analysis experiment request that is used for accepting the user, sets up decision-making constraint, object module and choice mechanism, and all kinds of schemes that the comparative analysis user submits to form the decision-making suggestion;
Communication layers MAS comprises: communication Agent;
Resource layer MAS comprises: material database Management Agent, KBM Agent, logical base Management Agent, example/countermeasure library management Agent, teledata are excavated Agent;
(6) said game subsystem, its function are to carry out intelligent man-machine game or the intelligent online game competitive power with the check decision-making;
Function layer MAS comprises: man-machine interface Agent, explanation Agent, manual intervention Agent;
Work layer MAS comprises: project Agent, operation Agent, Information Agent, natural language processing Agent, emulation Agent, graphic modeling Agent, game Agent, tactful Agent, target Agent; Wherein, game Agent is the game experiment request that is used for accepting the user, sets up judge's model, inference mechanism and game constraint, and the various schemes that the ruling user submits to form payoff; Strategy Agent is under game Agent control, carries out PROBLEM DECOMPOSITION according to projects Agent ability and state, and the bundle Task Distribution is given corresponding project Agent, and the sub-case of countermeasure of assembled item Agent feeds back to game Agent simultaneously; Target Agent is through game Agent acceptance problem and parameter, on LAN and wide area network, seeks online game target and strategy thereof and feeds back to game Agent;
Communication layers MAS comprises: communication Agent;
Resource layer MAS comprises: material database Management Agent, KBM Agent, logical base Management Agent, example/countermeasure library management Agent, teledata are excavated Agent;
(7) said intelligent decision simulating experimental system comprises material database, example/countermeasure storehouse, knowledge base, four types of local data bases of logical base; Material database is in order to storage and provide graphic modeling needed figure material; Knowledge base is in order to storage and the needed expertise of experiment, experimental rules, managerial knowledge and information data are provided; Logical base is in order to storage and the needed operation logic of experiment, inference logic, interpretation logic are provided; Example/countermeasure storehouse is in order to storage and Success in Experiment example/countermeasure and the needed same or similar case/countermeasure of experiment are provided.
2. according to the described intelligent decision simulating experimental system of claim 1, its related Agent comprises response type and mixed type, it is characterized in that:
(1) response type Agent: make corresponding action in response through the perception environmental change, represent Agent::=< Aid, P, A, see, action>with five-tuple, wherein, Aid is the sign of certain concrete Agent; A representes the behavior collection of Agent; P representes the visual state collection of Agent; See, action are used to portray inner observation process of Agent and behaviour decision making process;
(2) mixed type Agent: structure is divided into two-layer, and bottom is a responding layer, and symbolization is not represented and reasoning, can respond and handle the sudden variation of external environment condition fast, has higher priority; High-rise adopt that traditional artificial intelligence approach is planned, reasoning and decision-making, represent Agent::=< Aid, P, A with 11 tuples; R, Dva, Dar, Rule; See, next, estimate; Action >, wherein, R representes the set that all corresponding possible outcome states of all possible action schemes of Agent are constituted; Dva representes the rule of correspondence collection of visual state and scheme set; Corresponding relation between Dar representation scheme collection and the result phase; Rule is the set that concrete decision rule constitutes; Next is used to portray the inner thought process of Agent; Estimate representes selected scheme set, according to the rule of correspondence storehouse of scheme and result phase collection, confirms possible result phase; Other element implication is identical with response type Agent.
3. according to the described intelligent decision simulating experimental system of claim 2, have 4 layers of MAS, it is characterized in that:
(1) said each subsystem function layer MAS in order to fulfillment upper strata function, comprises following Agent:
Man-machine interface Agent: response type comprises input/output control module, communication module, behavior sensing module, I/O algorithm data-base and Agent state table and Agent dictionary;
Explain Agent: mixed type comprises and explains maker, conclusion control module, communication module, behavior sensing module, interpretation algorithms database and Agent state table and Agent dictionary;
Manual intervention Agent: response type comprises routine call control module, communication module, behavior sensing module, calls algorithm data-base;
Constraint Agent: response type comprises bound control module, communication module, behavior sensing module, bounding algorithm database and constraint condition table and Agent dictionary;
Reasoning Agent: i.e. inference machine, mixed type comprises problem deduce machine, regular control module, communication module, behavior sensing module, reasoning algorithm database and rule set and Agent dictionary;
(2) said each subsystem work layer MAS in order to realize system core business operation function, comprises following Agent:
Planning Agent: mixed type comprises mission planning device, decision-making and intelligent control module, communication module, behavior sensing module, planning algorithm database and Agent state table and Agent dictionary;
Assessment Agent: mixed type comprises evaluate alternatives module, assessment control module, communication module, behavior sensing module, assessment algorithm database and Agent state table and Agent dictionary;
Decision analysis Agent; Mixed type comprises scheme comparison module, decision analysis control module, communication module, behavior sensing module, decision making algorithm database and Agent state table and Agent dictionary;
Game Agent: mixed type comprises strategy generator, ruling and intelligent control module, communication module, behavior sensing module, game playing algorithm database and Agent state table and Agent dictionary;
Task Agent: mixed type comprises project planning device, decision-making and intelligent control module, communication module, behavior sensing module, planning algorithm database and Agent state table and Agent dictionary;
Target Agent: response type comprises target search control module, communication module, behavior sensing module, searching algorithm database and routing table and Agent dictionary;
Strategy Agent: mixed type comprises project planning device, decision-making and intelligent control module, communication module, behavior sensing module, planning algorithm database and Agent state table and Agent dictionary;
Project Agent: mixed type comprises work planning device, decision-making and intelligent control module, communication module, behavior sensing module, planning algorithm database and Agent state table and Agent dictionary;
Operation Agent: mixed type comprises the subproblem solver, finds the solution and intelligent control module, communication module, behavior sensing module, derivation algorithm database and Agent state table and Agent dictionary;
Information Agent: response type comprises information acquisition control module, communication module, behavior sensing module, gathers algorithm data-base and Agent state table and Agent dictionary;
Natural language processing Agent: mixed type comprises language processing module, information Recognition and intelligent control module, communication module, behavior sensing module, recognizer database, corpus and proprietary vocabulary and Agent dictionary;
Emulation Agent: response type comprises Simulation Control module, communication module, behavior sensing module, simulation algorithm database and function table and Agent dictionary;
Graphic modeling Agent: response type comprises graphic modeling control module, communication module, behavior sensing module, action algorithm data-base and function table and Agent dictionary;
(3) said each subsystem communication layer MAS has only communication Agent, and response type comprises the communication server, message transmission module, behavior sensing module, communication of algorithms database and registration table and routing table;
(4) said each subsystem resource layer MAS in order to realize the system bottom data management function, comprises following Agent:
The material database Management Agent: response type comprises the material classification and calls control module, communication module, behavior sensing module, classifies and call algorithm data-base and Agent state table and Agent dictionary;
KBM Agent: mixed type comprises knowledge comparison update module, knowledge classification and calls control module, knowledge-based inference module, communication module, behavior sensing module, comparison algorithm database, classifies and call algorithm data-base and Agent state table and Agent dictionary;
The logical base Management Agent: mixed type, comprise logic comparison update module, logical division and call control module, logic-based reasoning module, communication module, behavior sensing module, comparison algorithm database, classify and call algorithm data-base and Agent state table and Agent dictionary;
Example/countermeasure library management Agent: mixed type comprises example/countermeasure comparison update module, example/classification of game and calls control module, communication module, behavior sensing module, comparison algorithm database, classifies and call algorithm data-base and Agent state table and Agent dictionary;
Teledata is excavated Agent: mixed type comprises data evaluation and verification module, data mining control module, communication module, behavior sensing module, KDD component base and Internet resources retrieval and Agent dictionary.
4. according to the described intelligent decision simulating experimental system of claim 1, this system adopts 4 kinds of emulation experiment methods, it is characterized in that:
(1) mathematical model simulation is abstract and simplification with problem, sets up mathematical model, carries out behavior emulation through solution procedure, thereby predicts the outcome, and mathematical model simulation can combine with graphical simulation, reasoning from logic;
(2) X-Y scheme emulation is to utilize machine or artificial foundation, with dynamic planar graph emulation expression behaviour process, thus observation process and result of calculation;
(3) three-dimensional model emulation is to utilize machine or artificial foundation, with the comprehensive expression behaviour process of dynamic 3 D graphics emulation, thus observation process and result of calculation;
(4) reasoning from logic emulation is to give expertise with problem and environmental parameter, utilizes inference mechanism and algorithm predicts process and result, combines with mathematical model simulation usually.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2009100130451A CN101615265B (en) | 2009-08-11 | 2009-08-11 | Intelligent decision simulating experimental system based on multi-Agent technology |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2009100130451A CN101615265B (en) | 2009-08-11 | 2009-08-11 | Intelligent decision simulating experimental system based on multi-Agent technology |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101615265A CN101615265A (en) | 2009-12-30 |
| CN101615265B true CN101615265B (en) | 2012-07-04 |
Family
ID=41494893
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2009100130451A Expired - Fee Related CN101615265B (en) | 2009-08-11 | 2009-08-11 | Intelligent decision simulating experimental system based on multi-Agent technology |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101615265B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| RU2665045C2 (en) * | 2015-10-29 | 2018-08-27 | Акционерное общество "Российская корпорация ракетно-космического приборостроения и информационных систем" (АО "Российские космические системы") | System for modeling situations relating to conflicts and/or competition |
| CN112433608A (en) * | 2020-11-19 | 2021-03-02 | 北京航空航天大学 | Automatic identification method for human-computer information interaction risk scene |
Families Citing this family (35)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101782976B (en) * | 2010-01-15 | 2013-04-10 | 南京邮电大学 | Automatic selection method for machine learning in cloud computing environment |
| CN101958917B (en) * | 2010-03-24 | 2013-02-06 | 北京航空航天大学 | Cloud manufacturing system-oriented method for measuring and enhancing flexibility of resource service composition |
| CN101944113B (en) * | 2010-09-16 | 2013-05-15 | 华中科技大学 | Cloud computing system-based data acquisition system |
| CN102468975B (en) * | 2010-11-16 | 2014-09-10 | 苏州搜能信息科技有限公司 | Resource management simulation cloud computing system for cloud computing mobile network and application system thereof |
| CN102207928B (en) * | 2011-06-02 | 2013-04-24 | 河海大学常州校区 | Reinforcement learning-based multi-Agent sewage treatment decision support system |
| CN102508995A (en) * | 2011-09-26 | 2012-06-20 | 河南理工大学 | Coal mine accident simulating method and system based on multi-intelligent agent |
| CN102368783B (en) * | 2011-10-14 | 2014-02-05 | 深圳市京华科讯科技有限公司 | Cloud equipment macro control method |
| CN102420860A (en) * | 2011-11-27 | 2012-04-18 | 姜晓越 | Method and system for asynchronously providing solution service for submitted problem |
| CN102779204A (en) * | 2012-06-15 | 2012-11-14 | 北京理工大学 | Client-service-oriented multinational game strategic decision simulation and deduction high level architecture (HLA) simulation system |
| CN103336442B (en) * | 2013-07-04 | 2016-03-16 | 南昌航空大学 | A kind of aircraft electrical power system Hardware In The Loop Simulation Method based on AGENT modeling technique |
| CN103399962A (en) * | 2013-08-24 | 2013-11-20 | 河南星智发明电子科技有限公司 | Intelligent addition invention system |
| CN103530400A (en) * | 2013-10-23 | 2014-01-22 | 合山市科学技术情报研究所 | Information recognition device |
| CN103605695A (en) * | 2013-11-05 | 2014-02-26 | 佛山职业技术学院 | Internet based artificial-intelligence knowledge logic system and method thereof |
| CN104144078A (en) * | 2014-07-11 | 2014-11-12 | 北京哈工大计算机网络与信息安全技术研究中心 | Network safety experimental result management method and system |
| CN104199971B (en) * | 2014-09-23 | 2017-06-16 | 北京军石科技有限公司 | Visualization intelligence analysis method and system based on standard knowledge framework |
| CN104392248A (en) * | 2014-11-10 | 2015-03-04 | 苏州乐聚一堂电子科技有限公司 | Personifying memory type interaction artificial intelligence expert system |
| CN108229685B (en) * | 2016-12-14 | 2021-11-02 | 中国航空工业集团公司西安航空计算技术研究所 | Air-ground integrated unmanned intelligent decision-making method |
| CN106845795B (en) * | 2016-12-27 | 2019-12-27 | 合肥城市云数据中心股份有限公司 | Business transfer system based on multi-Agent technology and business transfer method thereof |
| CN106709017B (en) * | 2016-12-27 | 2018-01-26 | 山东麦港数据系统有限公司 | A kind of aid decision-making method based on big data |
| CN108960880A (en) * | 2017-05-23 | 2018-12-07 | 王四春 | A kind of cross-border electric business commercial affairs big data Analysis of Policy Making and data processing method |
| CN107133416B (en) * | 2017-05-24 | 2020-02-14 | 西北工业大学 | UUV multilayer hybrid immune intelligent body structure modeling method |
| CN107122842A (en) * | 2017-06-21 | 2017-09-01 | 苏州发飚智能科技有限公司 | The issue of bidding documents that a kind of Multi -Agent is supported should mark method and system of booking rooms |
| CN108846570B (en) * | 2018-06-08 | 2022-02-01 | 武汉理工大学 | Method for solving resource-limited project scheduling problem |
| CN109086523B (en) * | 2018-08-02 | 2022-11-11 | 湘潭大学 | Automatic generation method of power supply design experiment questions based on cognitive computation model |
| CN109472363B (en) * | 2018-10-29 | 2021-11-23 | 潘颖慧 | Interpretable competitor modeling method |
| CN109522356B (en) * | 2018-11-13 | 2022-03-11 | 中国核动力研究设计院 | Nuclear reactor digital experiment system |
| CN109636012A (en) * | 2018-11-26 | 2019-04-16 | 西南电子技术研究所（中国电子科技集团公司第十研究所） | Intelligence sentences the processing system that card predicted events occur |
| CN111401590A (en) * | 2018-12-14 | 2020-07-10 | 中国航天系统工程有限公司 | Prediction method based on multi-Agent forest ecosystem |
| US20220026884A1 (en) * | 2019-01-25 | 2022-01-27 | Siemens Aktiengesellschaft | Autonomous coordination of devices in industrial environments |
| CN109901821B (en) * | 2019-03-08 | 2022-07-22 | 北京工业大学 | PASP-based multi-Agent belief coordination method |
| CN110210838B (en) * | 2019-06-05 | 2022-08-26 | 南京邮电大学 | Government affair system design method and government affair system |
| CN111601277B (en) * | 2020-04-26 | 2022-09-13 | 北京踏歌智行科技有限公司 | On-line simulation system for unmanned mining operation |
| CN112247962B (en) * | 2020-10-19 | 2021-10-08 | 中国科学技术大学 | Man-machine game control method and system for upper limb wearable robot |
| CN112581085B (en) * | 2020-12-15 | 2024-01-05 | 北京动力机械研究所 | Visual flow management method and device for engine design |
| CN113947377B (en) * | 2021-10-22 | 2023-05-30 | 浙江正泰仪器仪表有限责任公司 | Laboratory management system |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1912837A (en) * | 2006-08-31 | 2007-02-14 | 上海交通大学 | Expansable distributed system of supporting large scale micro-traffic simulation |
| CN101477581A (en) * | 2008-12-19 | 2009-07-08 | 上海理工大学 | Multi-agent area road intersection signal integrated control simulation system |
-
2009
- 2009-08-11 CN CN2009100130451A patent/CN101615265B/en not_active Expired - Fee Related
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1912837A (en) * | 2006-08-31 | 2007-02-14 | 上海交通大学 | Expansable distributed system of supporting large scale micro-traffic simulation |
| CN101477581A (en) * | 2008-12-19 | 2009-07-08 | 上海理工大学 | Multi-agent area road intersection signal integrated control simulation system |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| RU2665045C2 (en) * | 2015-10-29 | 2018-08-27 | Акционерное общество "Российская корпорация ракетно-космического приборостроения и информационных систем" (АО "Российские космические системы") | System for modeling situations relating to conflicts and/or competition |
| CN112433608A (en) * | 2020-11-19 | 2021-03-02 | 北京航空航天大学 | Automatic identification method for human-computer information interaction risk scene |
| CN112433608B (en) * | 2020-11-19 | 2022-04-12 | 北京航空航天大学 | Automatic identification method for human-computer information interaction risk scene |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101615265A (en) | 2009-12-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101615265B (en) | Intelligent decision simulating experimental system based on multi-Agent technology | |
| Clarkson et al. | ‘Signposting’, a parameter-driven task-based model of the design process | |
| Myers | CPEF: A continuous planning and execution framework | |
| Levis et al. | C4ISR architectures: I. Developing a process for C4ISR architecture design | |
| Xu et al. | Computer-aided process planning–A critical review of recent developments and future trends | |
| Morad et al. | Knowledge-based planning system | |
| Pillac et al. | On the dynamic technician routing and scheduling problem | |
| Ntuen et al. | Interface agents in complex systems | |
| CN108984956A (en) | Earth observation satellite architecture Design and optimization system | |
| Hajdasz | Flexible management of repetitive construction processes by an intelligent support system | |
| Rao et al. | An active intelligent decision support system—architecture and simulation | |
| Dunets et al. | Multi-agent system of IT project planning | |
| Villela et al. | The use of an enterprise ontology to support knowledge management in software development environments | |
| Kourtz | Artificial intelligence: a new tool for forest management | |
| Turban | Expert systems—another frontier for industrial engineering | |
| CN110210838B (en) | Government affair system design method and government affair system | |
| Burstein et al. | An approach to mixed-initiative management of heterogeneous software agent teams | |
| Karacal et al. | A formal structure for discrete event simulation. Part I: Modeling multiple level systems | |
| Turban et al. | The utilization of expert systems in OR/MS: an assessment | |
| Sawaragi et al. | An interactive system for modeling and decision support–Shinayakana system approach | |
| Skowron | Approximate reasoning in MAS: rough set approach | |
| Garetti et al. | Using neural networks for reactive scheduling | |
| Aleisa | Multilevel integration of simulation and facilities planning for large-scale systems | |
| Druzdzel | ESP: A mixed initiative decision-theoretic decision modeling system | |
| Popa | SPECIFIC PROCEDURES TO INCREASE EFFICIENCY OF THE DECISION-MAKING PROCESS IN THE CONTEXT OF THE VARIABLE GEOMETRY CONFLICT |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C57 | Notification of unclear or unknown address | ||
| DD01 | Delivery of document by public notice |
Addressee: Wang Liying Document name: Notification of Passing Preliminary Examination of the Application for Invention |
|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CB03 | Change of inventor or designer information |
Inventor after: Xie Rui Inventor after: Lu Jun Inventor after: Wang Liying Inventor before: Lu Jun Inventor before: Wang Liying |
|
| CB03 | Change of inventor or designer information | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20170321 Address after: 300191 Tianjin city Nankai District wangdingdi Fung Yuen Building 6 South Lane 3 door 601 Patentee after: Xie Rui Address before: Huangpu Qixian garden in Ganjingzi District of Dalian City Road 116023 Liaoning province No. 15 Building 1 unit 501 Patentee before: Lu Jun |
|
| TR01 | Transfer of patent right | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20120704 Termination date: 20180811 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |