CN106779583A - Extra large frock based on artificial neural network is for project work STRUCTURE DECOMPOSITION System and method for - Google Patents
Extra large frock based on artificial neural network is for project work STRUCTURE DECOMPOSITION System and method for Download PDFInfo
- Publication number
- CN106779583A CN106779583A CN201611055784.3A CN201611055784A CN106779583A CN 106779583 A CN106779583 A CN 106779583A CN 201611055784 A CN201611055784 A CN 201611055784A CN 106779583 A CN106779583 A CN 106779583A
- Authority
- CN
- China
- Prior art keywords
- project
- neural network
- work
- artificial neural
- extra large
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/10—Office automation; Time management
- G06Q10/103—Workflow collaboration or project management
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- Entrepreneurship & Innovation (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Economics (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention is to provide a kind of extra large frock based on artificial neural network for project work STRUCTURE DECOMPOSITION system and method.Including OBS information modules, item view module, job analysis module and system management module;The OBS information modules are the specific organizational unit for describing responsible each project activity;The item view module is used for addition and the history item data storage of project information;The job analysis module is used to decompose project work structure;The system management module, for the management service of system.The present invention solves the problems, such as that sea work enterprise work STRUCTURE DECOMPOSITION takes time and effort at present, greatly reduce the error that workload and knowledge blind area are brought, also the execution and management to project is had laid a good foundation, more excellent decomposition texture makes shipyard improve production efficiency, increase economic efficiency, improve product quality.
Description
Technical field
System and method of the extra large frock for project work STRUCTURE DECOMPOSITION is realized by computer the present invention relates to a kind of,
Specifically a kind of extra large frock based on artificial neural network for project work STRUCTURE DECOMPOSITION System and method for.
Background technology
Artificial neural network ANN (Artificial Neural Network) is one kind of artificial intelligence, and it is from physiology
Angle simulate the mankind intelligency activity, the brain structure and function of the mankind are studied with the method for mathematics or physics.Profit
The general thought of employment artificial neural networks solve problem is the method for the learning knowledge for simulating the mankind, using training sample to nerve
Network is trained, the ability for making it have " memory " or even " reasoning " to certain class knowledge.Artificial neural network can solve multiple
Miscellaneous problem of nonlinear mapping, and with good fault-tolerance and adaptivity, therefore it is widely used in pattern-recognition, number
According to fields such as excavation, robot control, expert systems.
The project plan is key link in project management process, and whether plan pay has influence on project subsequent implementation and enter
Exhibition.Research for the project plan has a lot, and used as the basis of the project plan, project work STRUCTURE DECOMPOSITION should be closed by emphasis
Note, but research focuses more on activity sequence and resource allocation based on network plan method and project management techniques at present
The problems such as.
Work structure decomposition process is the process of identification and analysis project deliverable achievement and related work.For project work
Make the proposition of project work STRUCTURE DECOMPOSITION method and criterion that the achievement in research of STRUCTURE DECOMPOSITION is focusing more under specific condition, with
And the foundation of project work structure decomposition model.These researchs are only actual items work structure decomposition and provide theoretical foundation,
Directive function for the work of actual items work structure decomposition is limited.
Complex Sea frock for project work structure compared with ordinary skill project in complexity, uncertain and amount of activity
On all greatly increase.General project work STRUCTURE DECOMPOSITION method is for extra large work equipment Project and does not apply to.For Complex Sea work
Equipment Project work structure decomposition demand, it is a set of based on artificial neural network only from decomposition method and realization rate on
Extra large frock can well solve Complex Sea frock for project work STRUCTURE DECOMPOSITION for project work STRUCTURE DECOMPOSITION accessory system
Problem.
The content of the invention
It is an object of the invention to provide it is a kind of can improve shipbuilding efficiency, improve product quality based on artificial god
Through the extra large frock of network for project work STRUCTURE DECOMPOSITION system.It is a kind of based on ANN the present invention also aims to provide
The extra large frock of network is for project work STRUCTURE DECOMPOSITION method.
Extra large frock based on artificial neural network of the invention includes OBS for project work STRUCTURE DECOMPOSITION system
(Organizational Breakdown Structure, obs) information module, item view module, work point
Solution module and system management module;
The OBS information modules be for describe be responsible for each project activity specific organizational unit, including OBS addition,
The Limitation on Liability is divided, and operation, saddlebag and relevant departments or unit are connected, and ultimately forms OBS structures and project
Internal data is mutually butted;
The item view module is used for addition and the history item data storage of project information, including project is added, gone through
History project data is imported, item characteristic parameter editor, can be added and be checked the title of project, classification society, initial time, go into operation
The characteristic parameter information of time, priority, owner information and project is entered edlin and is checked, simultaneously for enterprise's history entries
Mesh number is according to lead-in item view;
The job analysis module is used to decompose project work structure, including items selection, artificial neural network
Training and application artificial neural network the intelligence of new projects' work structuring is decomposed, selection needs the item of work structure decomposition
Mesh, the decomposition using artificial neural network to new projects' work structuring is decomposed the data for completing and enters historical data base, improves item
Mesh numeric field data, while carrying out regular exercise and maintenance to artificial neural network structure;
The system management module, for the management service of system.
Extra large frock based on artificial neural network of the invention includes for project work STRUCTURE DECOMPOSITION method:
Step one, is matrix form by project work thaumatropy, sets up stratification work structuring element model;
Step 2, sets up extra large frock for item domains, covers the item number of item characteristic parameter and project work structural element
According to;
Step 3, design is applied to delamination modularization artificial neural network knot of the extra large frock for project work STRUCTURE DECOMPOSITION
Structure, and typical sea frock is applied to for project work STRUCTURE DECOMPOSITION, by the training and test of project numeric field data, stablized
Neural network structure, sets up the incidence relation of item characteristic parameter and project work structure;
Step 4, the neural network structure based on step 3 realizes project work to a work structure decomposition for new projects
Make the intelligence decomposition of structure, obtain the complete work structuring of new projects.
Extra large frock based on artificial neural network of the invention for project work STRUCTURE DECOMPOSITION method and system, to solve mesh
The problem that preceding extra large work Enterprise Project work structure decomposition takes time and effort is target.Artificial neural network technology is applied to extra large frock
Standby project work STRUCTURE DECOMPOSITION problem, extra large frock is realized for the huge task amount of project work STRUCTURE DECOMPOSITION by equivalency transform
Efficient process, while realize extra large frock using historical data and artificial neural network decomposed for the intelligence of project work structure,
In conjunction with information-based means, the above method is encapsulated into extra large frock for project work STRUCTURE DECOMPOSITION accessory system, solved at present
The problem that extra large work enterprise work STRUCTURE DECOMPOSITION takes time and effort, greatly reduces the error that workload and knowledge blind area are brought, and also gives
The execution and management of project are had laid a good foundation, and more excellent decomposition texture makes shipyard improve production efficiency, improve economic effect
Benefit, improve product quality.
Brief description of the drawings
Fig. 1 is extra large frock of the invention for project levelization work structural element model schematic;
Fig. 2 is extra large frock of the invention for item domains schematic diagram;
Fig. 3 is delamination modularization artificial neural network structure schematic diagram of the invention;
Fig. 4 is that the extra large frock based on artificial neural network of the invention is illustrated for project work STRUCTURE DECOMPOSITION method flow
Figure;
Fig. 5 is that the extra large frock based on artificial neural network of the invention is shown for project work STRUCTURE DECOMPOSITION systematic functional structrue
It is intended to.
Specific embodiment
Extra large frock based on artificial neural network of the invention includes for project work STRUCTURE DECOMPOSITION system:OBS information moulds
Block, item view module, job analysis module and system management module.OBS information modules, for describing to be responsible for the work of each project
Dynamic specific organizational unit, including OBS additions, the Limitation on Liability such as divide at the function, by operation item, saddlebag etc. and relevant departments or
Person's unit by different level, is systematically connected, and carries out to be carried out when project work STRUCTURE DECOMPOSITION is operated in project engineer
Person liable selects, and ultimately forms being mutually butted for OBS structures and project internal data;Item view module, for project information
Addition and history item data storage, the work(such as including project addition, the importing of history item data, item characteristic parameter editor
Can, the outline information such as title, classification society, initial time, on-stream time, priority, person liable of project can be added and check,
And the characteristic parameter information of project is entered edlin and is checked, can be regarded with lead-in item simultaneously for enterprise's history item data
Figure;Job analysis module, for being decomposed to project work structure, including items selection, artificial neural network training and
The intelligence of new projects' work structuring is decomposed using artificial neural network, selection needs the project of work structure decomposition, using people
Decomposition of the artificial neural networks to new projects' work structuring, decomposes the data for completing and enters historical data base, improves project numeric field data,
Regular exercise and maintenance can be carried out to artificial neural network structure simultaneously, it is ensured that artificial neural network structure's parameter is in optimal
State, the validity of safeguards system function and the optimization of project work STRUCTURE DECOMPOSITION result;System management module, for system
Management service.
Extra large frock based on artificial neural network of the invention includes for project work STRUCTURE DECOMPOSITION method:According to extra large frock
Standby project work STRUCTURE DECOMPOSITION feature, is matrix form by project work thaumatropy, sets up stratification work structuring element mould
Type;Extra large frock is set up for item domains, covers the project data such as item characteristic parameter and project work structural element;Design is applied to
Extra large frock and is applied to typical sea work equipment Project for the delamination modularization artificial neural network structure of project work STRUCTURE DECOMPOSITION
Work structure decomposition, by the training and test of project numeric field data, the neural network structure stablized sets up item characteristic ginseng
The incidence relation of number and project work structure;Based on the neural network structure to a work structure decomposition for new projects, realize
The intelligence decomposition of project work structure, obtains the complete work structuring of new projects.
The invention will be further described for citing below in conjunction with the accompanying drawings.
The species of extra large work equipment Project is varied, such as all kinds of drilling platforms projects, production platform project, Floating Storage
Ship, crane ship, pipe laying barge, diving boat etc., various extra large work equipment Projects can be entered using the method for the present invention and system
Row work structure decomposition, and polytype extra large work equipment Project can be classified in system, realize that specific extra large frock is standby
The intelligence decomposition of project work structure, below by taking the work equipment Project of a certain typical case sea as an example.
Worked structural element model schematic for project levelization with reference to the extra large frock of Fig. 1, the model includes:101 typical cases
Extra large frock is for project work decomposition texture, 102 equivalent adjacency matrix, 103 incidence matrix.
Typical sea frock lists part work breakdown structure (WBS) for project work decomposition texture 101.
Work breakdown structure (WBS) shown in 101 is described using directed acyclic graph G=(V, E), V is node set, represented
The node elements of Project decomposition structure;E is the set on side, represents the exploded relationship between node.
The digraph can be represented that the ranks of adjacency matrix are all the items at different levels of decomposition completion by the equivalent adjacency matrix A of G
Mesh saddlebag, if the continuation of row structure is decomposed includes array structure, the element of the adjacency matrix position is 1, is otherwise 0.
The directed acyclic graph and equivalent adjacency matrix of the dotted line frame inner structure of work breakdown structure (WBS) 101 are as shown in Figure 102.
Extra large frock is split as having for referable thing and both operations performed to it for the saddlebag of project work structure
Machine is combined.
This marriage relation is expressed with incidence matrix, is row coordinate with all referable things that project is included, with project bag
The all process steps for containing are row coordinate, and incidence matrix element represents that the project work structure is combined to form comprising the position ranks for 1
Saddlebag, element is the saddlebag that the 0 expression project work structure is combined to form not comprising the position ranks.
The referable thing of pile leg structure includes pipeline, sacrificial anode, guide-rail plate, water word in work breakdown structure (WBS) 101;Operation
Including prefabricated, installation, welding, pressure testing, then incidence matrix is shown in 103.
With reference to Fig. 2 extra large frock for item domains schematic diagram, including:201 item characteristic parameters, 202 project referable things,
203 project operations, 204 project work structure adjacency matrix and 205 project work structure connection matrix datas.
201 define item characteristic parameter, it is assumed that it is Tp, item that the characteristic parameter of extra large work equipment Project p integrates the element for including
Comprising n similar project in mesh domain, then the characteristic parameter included in item domains can be write as MT.
Tp is the subset of MT, so the characteristic parameter of any project p can find in MT.
202 define project referable thing, and 203 define project operation, it is assumed that the job analysis knot of extra large work equipment Project P
Structure can be represented with the element set Fp of the element set Dp of referable entity and operation, comprising n similar project in item domains, that
The referable thing and operation included in item domains can be write as MD and MF.
Dp and Fp are respectively the subsets of MD and MF, so the referable entity elements and operation of any project P can be
Found in item domains.
204 define item domains incidence matrix, and using the entity elements in item domains as longitudinal amount, operation is used as transverse direction
Amount, the element in each crosspoint as MR-Matrix element, you can obtain the matrix of MR-Matrix.
Similarly for project P, its corresponding R-Matrix is the submatrix of MR-Matrix, therefore, MR- in item domains
Matrix contains all possible operations of project P with the incidence relation for decomposing entity.
205 define item domains adjacency matrix, obtain after MR-Matrix, and each element is work point in order matrix
The element of a certain ranks in the equivalent adjacency matrix of solution structure, you can obtain item domains adjacency matrix MA.
Exploded relationship between the work breakdown structure (WBS) element of any project P can find in MA.
With reference to Fig. 3, after defining stratification work structuring element model and extra large frock for item domains, design is the present invention close
The artificial neural network structure of key.
Because the three first layers of extra large work equipment Project have its fixed structure, the difficult point during work structure decomposition is from the 3rd
Layer resolves into saddlebag one by one to the 4th layer of decomposition by specialty, therefore the foundation of Module Division is that project work is included
Each specialty.
Design layering artificial neural network structure, because extra large frock is huge for project work number of structures, by artificial
Neural network structure is layered, wherein 301 layers include two submodules of ANND and ANNF, 302 layers include an ANNR module,
303 layers include an ANNA modules
The input of 301 layers of ANND and ANNF modules is item characteristic parameter vector, and output is project referable entity and work
Order elements collection, its effect is the referable entity and operation that project of finding out is included.
The input of 302 layers of ANNR modules is referable entity elements collection, operation element set and item characteristic parameter vector,
Output is work structuring incidence matrix, and its effect is to find out all working structure node element;
In the incidence matrix that ANNR outputs are obtained, element value is the serial number between 0 and 1, as ANNA
, it is necessary to convert before the input of module, first conversion is that serial number is changed into 0 or 1 with critical value parameter, is more than
Critical value is assigned to 1, less than critical value is assigned to 0;
Because the input and output of artificial nerve network model all should be the forms of vector, second conversion is to associate
Matrix changes into the form of vector, by Factorization algorithm into column vector one by one during conversion.Then will conversion gained vector and project
Input of the characteristic parameter vector together as ANNA modules.
The input of 303 layers of ANNA modules is work structuring, and output is work structuring adjacency matrix.
With reference to Fig. 4 the extra large frock based on artificial neural network for project work STRUCTURE DECOMPOSITION method flow schematic diagram, bag
Include:S401 project datas prepare, S402 initialization neural network parameter, S403 neural metwork trainings, S404 neutral nets are tested
With S405 Application of Neural Network.
S401 project datas prepare:For the problem to be solved, suitable item domains are selected.Included in historical data base
Data not only include item domains mean terms purpose three first layers element set, referable entity elements collection, operation element set, WBS association squares
Battle array and adjacency matrix, also including item domains characteristic parameter collection, these data will be when neural network model be set up as training
And test data.
The characteristic parameter of project is built into parameter vector, is represented with vector T.
Due to character may be included in the characteristic parameter of project, in this step with unified rule by character style
Characteristic parameter coding, then by encoded translated into binary representation, such as characteristic parameter of certain extra large work equipment Project
In have " self-elevating drilling platform ", in the application we by " self-elevating drilling platform " Uniform provisions into coding " 001 ", then with
Binary form represents " 001 ", and the element of the final item characteristic parameter vector that we obtain is binary number 0 or 1.
S402 initializes neural network parameter:Needed for setting the initial connection weight and threshold value, and module of all modules
Target accuracy rate.
S403 neural metwork trainings:Using multilayer perceptron neutral net, back-propagation algorithm is used as training algorithm
Training sample is trained to neutral net, by the connection weight for training revision neutral net.
S404 neutral nets are tested:Performance test sample is tested neutral net.
The characteristic parameter vector of input test sample, obtains the WBS matrixes of project, then by test result and actual WBS
Result is compared, the accuracy rate of result of calculation.
The accuracy rate of each module and the accuracy rate of whole neural network structure are calculated during test respectively.
Neural network model is used for new projects' work structure decomposition by S405:In the neural network model that training is optimized
Afterwards, new projects' work structure decomposition is built with neutral net, the characteristic parameter of new projects is indicated with vector, as mould
The input of type.
The result that model is calculated is indicated with adjacency matrix, invalid operation therein is deleted, that is, obtained
This project work STRUCTURE DECOMPOSITION structure chart.
Show for project work STRUCTURE DECOMPOSITION accessory system functional structure with reference to the extra large frock based on artificial neural network of Fig. 5
It is intended to, system function module includes module 501OBS information, the item view of module 502, the job analysis of module 503 and module 504
System administration.
Module 501 is used for the specific organizational unit for describing to be responsible for each project activity, including OBS additions, the Limitation on Liability are drawn
Grade function;
Operation item, saddlebag etc. and relevant departments or unit by different level, are systematically connected, in plan engineering
Shi Jinhang project works STRUCTURE DECOMPOSITION can carry out person liable's selection when operating, and ultimately form OBS structures and project internal data
Be mutually butted.
Module 502 is used for addition and the history item data storage of project information, including project addition, history item data
Importing, item characteristic parameter edition function;
Add and check that the outlines such as title, classification society, initial time, on-stream time, priority, the person liable of project are believed
Breath, and the characteristic parameter information of project is entered edlin and is checked, can be with lead-in item simultaneously for enterprise's history item data
View.
Module 503 be used for project work structure is decomposed, including items selection, artificial neural network training and should
Employment artificial neural networks are decomposed to the intelligence of new projects' work structuring;
Selection needs the project of work structure decomposition, and the decomposition using artificial neural network to new projects' work structuring divides
Solve the data for completing and enter historical data base, project numeric field data is improved, while can be carried out periodically to artificial neural network structure
Training and maintenance, it is ensured that artificial neural network structure's parameter is in optimum state, the validity and project work of safeguards system function
Make the optimization of structure decomposition result.
Module 504 is used for the management service of system, including user management and rights management.
Knowable to foregoing description, the extra large frock based on artificial neural network for project work STRUCTURE DECOMPOSITION method and system,
Artificial neural network technology is applied to extra large frock for project work STRUCTURE DECOMPOSITION problem, extra large frock is realized by equivalency transform
The efficient process of the standby project work huge task amount of STRUCTURE DECOMPOSITION, while realizing sea using historical data and artificial neural network
Frock is decomposed for the intelligence of project work structure, in conjunction with information-based means, the above method is encapsulated into extra large work equipment Project work
Make STRUCTURE DECOMPOSITION accessory system, solve the problems, such as that sea work enterprise work STRUCTURE DECOMPOSITION takes time and effort at present, greatly reduces work
The error that work amount and knowledge blind area are brought, the also execution and management to project is had laid a good foundation, more excellent decomposition texture
Make shipyard improve production efficiency, increase economic efficiency, improve product quality.
The preferred embodiments of the present invention are the foregoing is only, the present invention is not restricted to, for those skilled in the art
For, the present invention can have various modifications and variations.All any modifications within the spirit and principles in the present invention, made, etc.
With replacement, improvement etc., should be included within the scope of the present invention.
Claims (10)
1. a kind of extra large frock based on artificial neural network is for project work STRUCTURE DECOMPOSITION system, it is characterized in that:Including OBS information
Module, item view module, job analysis module and system management module;The OBS information modules are for describing to be responsible for every
The specific organizational unit of individual project activity;The item view module is used for the addition and the storage of history item data of project information
Deposit;The job analysis module is used to decompose project work structure;The system management module, for the management of system
Safeguard.
2. the extra large frock based on artificial neural network according to claim 1 is for project work STRUCTURE DECOMPOSITION system, and it is special
Levying is:The OBS information modules include that OBS additions, the Limitation on Liability are divided, by operation, saddlebag and relevant departments or list
Position connects, and ultimately forms being mutually butted for OBS structures and project internal data.
3. the extra large frock based on artificial neural network according to claim 1 is for project work STRUCTURE DECOMPOSITION system, and it is special
Levying is:The item view module includes project addition, the importing of history item data, item characteristic parameter editor, can add
The characteristic parameter of title, classification society, initial time, on-stream time, priority, owner information and project with the project of checking
Information is entered edlin and is checked, simultaneously for enterprise's history item data lead-in item view.
4. the extra large frock based on artificial neural network according to claim 1 is for project work STRUCTURE DECOMPOSITION system, and it is special
Levying is:The job analysis module includes items selection, the training of artificial neural network and application artificial neural network to new item
The intelligence decomposition of mesh work structuring, selection needs the project of work structure decomposition, new projects are worked using artificial neural network
The decomposition of structure, decomposes the data for completing and enters historical data base, project numeric field data is improved, while to artificial neural network structure
Carry out regular exercise and maintenance.
5. a kind of extra large frock based on artificial neural network is for project work STRUCTURE DECOMPOSITION method, it is characterized in that:
Step one, is matrix form by project work thaumatropy, sets up stratification work structuring element model;
Step 2, sets up extra large frock for item domains, covers the project data of item characteristic parameter and project work structural element;
Step 3, design is applied to delamination modularization artificial neural network structure of the extra large frock for project work STRUCTURE DECOMPOSITION, and
Typical sea frock is applied to for project work STRUCTURE DECOMPOSITION, by the training and test of project numeric field data, the nerve stablized
Network structure, sets up the incidence relation of item characteristic parameter and project work structure;
Step 4, the neural network structure based on step 3 realizes project work knot to a work structure decomposition for new projects
The intelligence decomposition of structure, obtains the complete work structuring of new projects.
6. the extra large frock based on artificial neural network according to claim 5 is for project work STRUCTURE DECOMPOSITION method, and it is special
Levy is that step one is specifically included:
Extra large frock is equivalent to adjacency matrix for project work structure, the ranks of adjacency matrix are all the projects at different levels of decomposition completion
Saddlebag, is described using directed acyclic graph to work structuring, if the continuation of row structure is decomposed includes array structure, is abutted
The element of the matrix position is 1, is otherwise 0;
Extra large frock is split as organic knot of referable thing and both operations performed to it for the saddlebag of project work structure
Close, this marriage relation is expressed with incidence matrix, be row coordinate with all referable things that project is included, the institute included with project
There is operation for row coordinate, incidence matrix element is the work that the 1 expression project work structure is combined to form comprising the position ranks
Bag, element is the saddlebag that the 0 expression project work structure is combined to form not comprising the position ranks.
7. the extra large frock based on artificial neural network according to claim 5 is for project work STRUCTURE DECOMPOSITION method, and it is special
Levying is:Extra large frock includes that all items characteristic parameter, project referable thing, project operation, project work structure are adjacent for item domains
Connect matrix and project work structure connection matrix data.
8. the extra large frock based on artificial neural network according to claim 5 is for project work STRUCTURE DECOMPOSITION method, and it is special
Levy is that step 3 is specifically included:
Design artificial neural network structure, using item argument as input, the work structuring of project is realized being input to as output
The Nonlinear Processing process of output, realizes the association of item argument and project work structure;
Design modularized manual neural network structure, the foundation of Module Division is each specialty that project work is included;
Design layering artificial neural network structure, is layered to artificial neural network structure, wherein ground floor comprising ANND and
Two submodules of ANNF, the second layer includes an ANNR module, and third layer includes an ANNA module;
Data with extra large frock for item domains are trained and test to the artificial neural network structure for designing, and are stablized
Artificial neural network structure.
9. the extra large frock based on artificial neural network according to claim 8 is for project work STRUCTURE DECOMPOSITION method, and it is special
Levy is that the design layering artificial neural network structure specifically includes:
Each layer all includes 14 modules;
The input of ANND and ANNF modules is item characteristic parameter vector, and output is project referable entity and operation element set,
Its effect is the referable entity and operation that project of finding out is included;
The input of ANNR modules is referable entity elements collection, operation element set and item characteristic parameter vector, and output is work
Make structure incidence matrix, its effect is to find out all working structure node element;
The input of ANNA modules is work structuring, and output is work structuring adjacency matrix.
10. the extra large frock based on artificial neural network according to claim 9 is for project work STRUCTURE DECOMPOSITION method, and it is special
Levying is:In the incidence matrix that ANNR outputs are obtained, element value is the serial number between 0 and 1, as ANNA modules
Input before convert, first conversion is that serial number is changed into 0 or 1 with critical value parameter, more than critical value
1 is assigned to, 0 is assigned to less than critical value;
Second conversion is the form that incidence matrix is changed into vector, by Factorization algorithm into column vector one by one during conversion;
Then the input of gained vector and item characteristic parameter vector together as ANNA modules will be converted.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201611055784.3A CN106779583A (en) | 2016-11-25 | 2016-11-25 | Extra large frock based on artificial neural network is for project work STRUCTURE DECOMPOSITION System and method for |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201611055784.3A CN106779583A (en) | 2016-11-25 | 2016-11-25 | Extra large frock based on artificial neural network is for project work STRUCTURE DECOMPOSITION System and method for |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106779583A true CN106779583A (en) | 2017-05-31 |
Family
ID=58910731
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201611055784.3A Pending CN106779583A (en) | 2016-11-25 | 2016-11-25 | Extra large frock based on artificial neural network is for project work STRUCTURE DECOMPOSITION System and method for |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106779583A (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108170987A (en) * | 2018-01-23 | 2018-06-15 | 成都希盟科技有限公司 | The automatic hooking method of PBS structures based on BIM technology |
CN108170989A (en) * | 2018-01-23 | 2018-06-15 | 成都希盟科技有限公司 | Engineering construction model deriving method based on BIM technology |
CN109947898A (en) * | 2018-11-09 | 2019-06-28 | 中国电子科技集团公司第二十八研究所 | Based on intelligentized equipment failure test method |
CN112149042A (en) * | 2020-10-23 | 2020-12-29 | 青岛地铁集团有限公司运营分公司 | System and method for managing subway tunnel structure quality information in operation period |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101071412A (en) * | 2006-05-10 | 2007-11-14 | 何千军 | Neural network analysis system and method based on self-definition model |
CN103336991A (en) * | 2013-07-05 | 2013-10-02 | 广东电网公司电力科学研究院 | Resource allocation method and resource allocation system for line walking shortcut building project |
-
2016
- 2016-11-25 CN CN201611055784.3A patent/CN106779583A/en active Pending
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101071412A (en) * | 2006-05-10 | 2007-11-14 | 何千军 | Neural network analysis system and method based on self-definition model |
CN103336991A (en) * | 2013-07-05 | 2013-10-02 | 广东电网公司电力科学研究院 | Resource allocation method and resource allocation system for line walking shortcut building project |
Non-Patent Citations (1)
Title |
---|
SHARIFZADEH, M.等: "Designing work breakdown structure(WBS) in tunneling projects using neural networks", 《ARMA》 * |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108170987A (en) * | 2018-01-23 | 2018-06-15 | 成都希盟科技有限公司 | The automatic hooking method of PBS structures based on BIM technology |
CN108170989A (en) * | 2018-01-23 | 2018-06-15 | 成都希盟科技有限公司 | Engineering construction model deriving method based on BIM technology |
CN108170987B (en) * | 2018-01-23 | 2021-01-01 | 成都希盟泰克科技发展有限公司 | BIM technology-based PBS structure automatic hanging method |
CN108170989B (en) * | 2018-01-23 | 2021-03-02 | 成都希盟泰克科技发展有限公司 | Engineering construction model derivation method based on BIM technology |
CN109947898A (en) * | 2018-11-09 | 2019-06-28 | 中国电子科技集团公司第二十八研究所 | Based on intelligentized equipment failure test method |
CN109947898B (en) * | 2018-11-09 | 2021-03-05 | 中国电子科技集团公司第二十八研究所 | Equipment fault testing method based on intellectualization |
CN112149042A (en) * | 2020-10-23 | 2020-12-29 | 青岛地铁集团有限公司运营分公司 | System and method for managing subway tunnel structure quality information in operation period |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106779583A (en) | Extra large frock based on artificial neural network is for project work STRUCTURE DECOMPOSITION System and method for | |
Sahinidis | Optimization under uncertainty: state-of-the-art and opportunities | |
CN108595905A (en) | A kind of erosion failure quantitative forecasting technique based on BP neural network model | |
Andrews et al. | State of the art report on design methodology | |
CN105095961A (en) | Mixing system with artificial neural network and impulsive neural network | |
CN105095967A (en) | Multi-mode neural morphological network core | |
CN105095966A (en) | Hybrid computing system of artificial neural network and impulsive neural network | |
CN105095965A (en) | Hybrid communication method of artificial neural network and impulsive neural network | |
CN106096911A (en) | A kind of based role have people/unmanned collaborative command and control system and method thereof | |
Da et al. | Evaluation on connectivity of urban waterfront redevelopment under hesitant fuzzy linguistic environment | |
CN102646164A (en) | Land use change modeling method and system implemented in combination with spatial filtering | |
Jilg et al. | Optimized distributed control and topology design for hierarchically interconnected systems | |
CN112508192A (en) | Increment heap width learning system with degree of depth structure | |
CN113689128A (en) | Dynamic study and judgment and early warning system and method for ship navigation safety | |
Abd Rahman et al. | Flood mitigation measres using intuitionistic fuzzy dematel method | |
CN104656620A (en) | Comprehensive evaluation system for remanufacturing of heavy-duty machine tool | |
Hosen et al. | Navigating the Nexus of Economic Growth Global Warming and Business Innovation with AI | |
CN106021880A (en) | Jacket platform structure response computing method based on BP neural network | |
Iwańkowicz et al. | Self-classification of assembly database using evolutionary method | |
CN107644143A (en) | A kind of high-performance city CA model construction method based on vectorization and parallel computation | |
CN116402143A (en) | Intelligent ship index system construction and evaluation method | |
Taraska et al. | Review of assembly sequence planning methods in terms of their applicability in shipbuilding processes | |
Majstorović et al. | Industry 4.0 in Serbia-state of development | |
Wang et al. | Logistics forecast of malacca strait port using grey GM (1, N) model | |
CN114091883A (en) | Method, device, medium and equipment for predicting well leakage risk horizon before drilling |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170531 |