CN114036837B - Equipment combination method, system, equipment and storage medium based on co-construction sharing - Google Patents

Equipment combination method, system, equipment and storage medium based on co-construction sharing Download PDF

Info

Publication number
CN114036837B
CN114036837B CN202111315568.9A CN202111315568A CN114036837B CN 114036837 B CN114036837 B CN 114036837B CN 202111315568 A CN202111315568 A CN 202111315568A CN 114036837 B CN114036837 B CN 114036837B
Authority
CN
China
Prior art keywords
equipment
development
building
neighborhood
solution
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.)
Active
Application number
CN202111315568.9A
Other languages
Chinese (zh)
Other versions
CN114036837A (en
Inventor
豆亚杰
陈子夷
徐向前
刘泽水
鲁延京
谭跃进
杨克巍
姜江
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
National University of Defense Technology
Original Assignee
National University of Defense Technology
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by National University of Defense Technology filed Critical National University of Defense Technology
Priority to CN202111315568.9A priority Critical patent/CN114036837B/en
Publication of CN114036837A publication Critical patent/CN114036837A/en
Application granted granted Critical
Publication of CN114036837B publication Critical patent/CN114036837B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
    • 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
    • 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
    • G06N3/08—Learning methods
    • G06N3/082—Learning methods modifying the architecture, e.g. adding, deleting or silencing nodes or connections
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N5/00—Computing arrangements using knowledge-based models
    • G06N5/01—Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2111/00—Details relating to CAD techniques
    • G06F2111/04—Constraint-based CAD

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Evolutionary Computation (AREA)
  • Software Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Artificial Intelligence (AREA)
  • Data Mining & Analysis (AREA)
  • Computing Systems (AREA)
  • Mathematical Physics (AREA)
  • Computational Linguistics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Medical Informatics (AREA)
  • Computer Hardware Design (AREA)
  • Geometry (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The application relates to a method, a system, equipment and a storage medium for equipment combination based on co-building sharing. The method comprises the following steps: from a set of equipment to be developed, constructing a shared co-building combined model according to technical requirements and types of the development equipment and according to the number of alternative development units and equipment units to be received; aiming at the shared co-building combination model, designing DNNB & BH algorithm according to the deep neural network; carrying out overall association and calculation on the shared co-building combined model through the DNNB & BH algorithm to obtain a feasible solution of the shared co-building combined model; and according to the feasible solution, obtaining the specific schemes of equipment development task allocation, development time, development quantity and equipment combination selection. Compared with the traditional genetic algorithm and commercial optimizers widely applied in the market, the method has obvious speed advantages; the decision maker can be supported to conduct weapon equipment system top layer planning and decision.

Description

Equipment combination method, system, equipment and storage medium based on co-construction sharing
Technical Field
The present application relates to the field of military equipment, and in particular, to an equipment combining method, system, device and storage medium based on co-construction sharing.
Background
In contrast to general combinatorial planning problems and traditional equipment architecture development planning problems, complex equipment combinatorial optimization solution studies that consider co-building sharing face the following challenges and problems.
1) The simple combination of the conventional development planning and combination selection problems cannot guarantee "1+1=2". The problem of optimizing the combination of complex equipment is not only to consider the complexity of investment combinations in the co-construction stage, but also to consider the problem of function and resource sharing in a typical scene of an expected operation stage in advance, and the sharing in the co-construction stage and the operation stage in the investment research and development stage influences the optimization decision of the combination of complex equipment. Considering the co-building and sharing separately can solve the corresponding problems separately, but the optimization of each stage is not necessarily guaranteed to be global. This requires designing a corresponding optimization algorithm based on co-building shared double layer strategy to solve this problem.
2) The close association of the co-creation and sharing phases requires a stronger correlation between solutions. Considering that the combined optimization problem of the co-construction shared complex equipment is not simple mathematical calculation and is not the simulated simulation of cold ice, the construction and development of an equipment system in one stage are performed in a period of years or even more than ten years, and the research and development scheme in the co-construction stage and the distribution scheme in the sharing stage are not completely split, so that the related human factor effect is very important, and related historical data, expert experience, decision preference and the like are reflected on the combined planning scheme, so that the related inheritance and relevance among schemes in the stepwise development of one step are realized. How to represent this relevance when planning a solution is an important point in interactive decisions today.
3) The large number of constraints and associations gives the problem a higher complexity, requiring more efficient algorithms. The new combat concepts such as mosaic combat, distributed killing and multi-domain combat put forward higher requirements on future equipment system construction of the army, the new battlefield environment nowadays favors multifunctional, high-elasticity and high-adaptability complex equipment, the more complex equipment is associated with the increase of candidate schemes, and the complex equipment combination optimization problem is caused to appear in a combination explosion phenomenon along with the increase of the scale. Meanwhile, the problem of co-construction and sharing is comprehensively considered, so that the variable and constraint are increased, and the solution space is huge and complex.
Disclosure of Invention
In view of the foregoing, it is desirable to provide a method, a system, a computer device, and a storage medium for equipment combination based on co-building sharing.
In a first aspect, an embodiment of the present invention provides a method for combining equipment based on co-construction sharing, where the method includes:
From a set of equipment to be developed, constructing a shared co-building combined model according to technical requirements and types of the development equipment and according to the number of alternative development units and equipment units to be received;
aiming at the shared co-building combination model, designing DNNB & BH algorithm according to the deep neural network;
Carrying out overall association and calculation on the shared co-building combined model through the DNNB & BH algorithm to obtain a feasible solution of the shared co-building combined model;
and according to the feasible solution, obtaining the specific schemes of equipment development task allocation, development time, development quantity and equipment combination selection.
Further, the building a shared co-building combination model from the to-be-developed equipment set according to the technical requirements and types of the to-be-developed equipment and according to the number of the alternative development units and the to-be-received equipment units comprises:
According to equipment development planning, taking a co-building strategy as a drive, and selecting a plurality of pieces of equipment from equipment sets to be researched and developed to perform research and development work;
selecting a plurality of research and development units from the alternative research and development units to respectively bear different research and development tasks based on balanced meeting of different capacity requirements, and distributing research and development types of equipment according to technical requirements of research and development of the equipment;
According to the equipment cooperation requirement, the development expense budget, the constraint of the increase of the equipment amplitude quantity to the capacity improvement amplitude is taken as a basis, the sharing strategy is taken as a drive, and the shared co-building combination model based on the development time, the development expense and the capacity redundancy is designed.
Further, the performing overall association and calculation on the shared co-building combination model through the DNNB & BH algorithm to obtain a feasible solution of the shared co-building combination model, including:
Adopting DNNB & BH algorithm to comprehensively consider and synchronously solve and optimize the multi-stage equipment planning problem by means of a probability priority decision mode and a tree search exploration mode;
The inherent structure and the learning mechanism of the deep neural network are adopted, the self-relevance of the equipment combination scheme is obtained, the learning neural network model is utilized to assist the solving process, and the human factors are reflected in the solving process;
And accelerating the speed of approaching the optimal solution in the solution through the branch decision, trimming redundant branches through the delimitation decision, reducing the space-knowing scale, and obtaining the feasible solution of the shared co-building combined model.
Further, the performing overall association and calculation on the shared co-building combination model through the DNNB & BH algorithm to obtain a feasible solution of the shared co-building combination model, including:
obtaining a neighborhood solution set under different change strategies according to the feasible solutions given by the shared co-building combination model;
Inputting the solution matrix of the neighborhood solution set into respective branch decision depth neural network models, acquiring probability of converting the neighborhood change into the optimal solution, and taking the result with highest probability in the set obtained by the neighborhood change as a child node of a neighborhood structure branch;
Inputting the child node into a given decision depth neural network model again, determining whether to reserve the branch corresponding to the child node, and determining whether to repair the branch corresponding to the child node;
When the stopping requirement is met, selecting one of the sub-nodes in the search tree under the co-construction strategy, which has the highest corresponding probability, as the current solution under the co-construction strategy, and simultaneously selecting the corresponding sub-node in the search tree under the sharing strategy, and taking the current solution under the sharing strategy as the current solution.
On the other hand, the embodiment of the invention also provides a device combination system based on co-construction sharing, which comprises:
The shared co-building module is used for building a shared co-building combined model from a to-be-researched and developed equipment set according to the technical requirements and types of research and development equipment and according to the number of alternative research and development units and to-be-received equipment units;
The algorithm construction module is used for aiming at the shared co-construction combined model and designing DNNB & BH algorithm according to the deep neural network;
The model solving module is used for carrying out overall association and calculation on the shared co-built combined model through the DNNB & BH algorithm to obtain a feasible solution of the shared co-built combined model;
And the equipment combination module is used for obtaining specific schemes of equipment research and development task allocation, research and development time, research and development quantity and equipment combination selection according to the feasible solutions.
Further, the shared co-building module includes an equipment planning unit for:
According to equipment development planning, taking a co-building strategy as a drive, and selecting a plurality of pieces of equipment from equipment sets to be researched and developed to perform research and development work;
selecting a plurality of research and development units from the alternative research and development units to respectively bear different research and development tasks based on balanced meeting of different capacity requirements, and distributing research and development types of equipment according to technical requirements of research and development of the equipment;
According to the equipment cooperation requirement, the development expense budget, the constraint of the increase of the equipment amplitude quantity to the capacity improvement amplitude is taken as a basis, the sharing strategy is taken as a drive, and the shared co-building combination model based on the development time, the development expense and the capacity redundancy is designed.
Further, the model solving module includes a neural network auxiliary unit for:
Adopting DNNB & BH algorithm to comprehensively consider and synchronously solve and optimize the multi-stage equipment planning problem by means of a probability priority decision mode and a tree search exploration mode;
The inherent structure and the learning mechanism of the deep neural network are adopted, the self-relevance of the equipment combination scheme is obtained, the learning neural network model is utilized to assist the solving process, and the human factors are reflected in the solving process;
And accelerating the speed of approaching the optimal solution in the solution through the branch decision, trimming redundant branches through the delimitation decision, reducing the space-knowing scale, and obtaining the feasible solution of the shared co-building combined model.
Further, the model solving module comprises a branch-and-bound unit for:
obtaining a neighborhood solution set under different change strategies according to the feasible solutions given by the shared co-building combination model;
Inputting the solution matrix of the neighborhood solution set into respective branch decision depth neural network models, acquiring probability of converting the neighborhood change into the optimal solution, and taking the result with highest probability in the set obtained by the neighborhood change as a child node of a neighborhood structure branch;
Inputting the child node into a given decision depth neural network model again, determining whether to reserve the branch corresponding to the child node, and determining whether to repair the branch corresponding to the child node;
When the stopping requirement is met, selecting one of the sub-nodes in the search tree under the co-construction strategy, which has the highest corresponding probability, as the current solution under the co-construction strategy, and simultaneously selecting the corresponding sub-node in the search tree under the sharing strategy, and taking the current solution under the sharing strategy as the current solution.
The embodiment of the invention also provides computer equipment, which comprises a memory, a processor and a computer program stored on the memory and capable of running on the processor, wherein the processor realizes the following steps when executing the computer program:
From a set of equipment to be developed, constructing a shared co-building combined model according to technical requirements and types of the development equipment and according to the number of alternative development units and equipment units to be received;
aiming at the shared co-building combination model, designing DNNB & BH algorithm according to the deep neural network;
Carrying out overall association and calculation on the shared co-building combined model through the DNNB & BH algorithm to obtain a feasible solution of the shared co-building combined model;
and according to the feasible solution, obtaining the specific schemes of equipment development task allocation, development time, development quantity and equipment combination selection.
The embodiment of the invention also discloses a computer readable storage medium, which stores a computer program, the computer program realizing the following steps when being executed by a processor:
From a set of equipment to be developed, constructing a shared co-building combined model according to technical requirements and types of the development equipment and according to the number of alternative development units and equipment units to be received;
aiming at the shared co-building combination model, designing DNNB & BH algorithm according to the deep neural network;
Carrying out overall association and calculation on the shared co-building combined model through the DNNB & BH algorithm to obtain a feasible solution of the shared co-building combined model;
and according to the feasible solution, obtaining the specific schemes of equipment development task allocation, development time, development quantity and equipment combination selection.
The equipment combination method, system, computer equipment and storage medium based on co-construction sharing, wherein the method comprises the following steps: from a set of equipment to be developed, constructing a shared co-building combined model according to technical requirements and types of the development equipment and according to the number of alternative development units and equipment units to be received; aiming at the shared co-building combination model, designing DNNB & BH algorithm according to the deep neural network; carrying out overall association and calculation on the shared co-building combined model through the DNNB & BH algorithm to obtain a feasible solution of the shared co-building combined model; and according to the feasible solution, obtaining the specific schemes of equipment development task allocation, development time, development quantity and equipment combination selection. Under the drive of a co-construction shared double-layer strategy, the method can comprehensively consider the association between equipment schemes by utilizing DNNB & BH algorithm, explore solution spaces at different stages, quickly complete the convergence of the algorithm, and obtain the associated high-quality feasible solution of the complex equipment combination optimization problem under different targets in a short time, and has obvious speed advantage compared with the traditional genetic algorithm and commercial optimizers widely applied in the market; meanwhile, the resource efficient integration of research and development units in the co-construction stage is comprehensively considered, and different constraints such as resource sharing, mutual coordination and the like in an expected application scene in the sharing stage are considered, so that a decision maker can be supported to conduct weapon equipment system top-level planning and decision making, and a specific scheme comprising equipment research and development task allocation, research and development time, research and development quantity and equipment combination selection can be rapidly obtained.
Drawings
FIG. 1 is a flow diagram of an equipment combining method based on co-building sharing in one embodiment;
FIG. 2 is a flow diagram of building a shared co-building combined model in one embodiment;
FIG. 3 is a schematic diagram of a feasible solution flow for obtaining a shared co-building combination model through DNNB & BH algorithm in one embodiment;
FIG. 4 is a flow diagram of a DNNB & BH optimization algorithm solution flow in one embodiment;
FIG. 5 is a block diagram of an equipment assembly system based on co-construction sharing in one embodiment;
fig. 6 is an internal structural diagram of a computer device in one embodiment.
Detailed Description
The present application will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present application more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the application.
In one embodiment, as shown in fig. 1, there is provided an equipment combining method based on co-building sharing, the method comprising:
step 101, constructing a shared co-building combined model from a to-be-researched and developed equipment set according to technical requirements and types of research and development equipment and according to the number of alternative research and development units and to-be-received equipment units;
102, designing DNNB & BH algorithm according to the deep neural network aiming at the shared co-building combination model;
Step 103, carrying out overall association and calculation on the shared co-building combined model through the DNNB & BH algorithm to obtain a feasible solution of the shared co-building combined model;
And 104, obtaining specific schemes of equipment development task allocation, development time, development quantity and equipment combination selection according to the feasible solutions.
Specifically, for complex equipment combination optimization problems needing to be comprehensively considered and shared, the conventional equipment development planning solution method can calculate an optimal solution in a solution space of the problem, but it is noted that each possible equipment development planning solution corresponds to a completely independent solution space of the equipment combination selection problem, so that the solution of the equipment development planning problem and the equipment combination selection problem, which is only guaranteed to be the optimal solution of the development planning problem, is considered separately, the solution space of the equipment combination selection problem corresponding to the non-dominant solution of the rest of the development planning problems is not explored, and after the optimal development planning solution is obtained, a high-quality combination selection solution can be obtained. Therefore, the embodiment aims at the complex equipment combination optimization problem of the type, comprehensively considers the co-construction sharing strategy, and simultaneously explores the solution space of the problem under the common driving of the double-layer strategy, thereby realizing the comprehensive consideration and synchronous solution optimization of the related combination optimization problem.
Preferably, the method utilizes DNNB & BH algorithm under the drive of co-construction shared double-layer strategy to comprehensively consider the association between equipment schemes, explores solution spaces at different stages, rapidly completes the convergence of the algorithm, obtains the associated high-quality feasible solution of complex equipment combination optimization problem under different targets in a short time, and has obvious speed advantage compared with the traditional genetic algorithm and commercial optimizers widely applied in the market; meanwhile, the resource efficient integration of research and development units in the co-construction stage is comprehensively considered, and different constraints such as resource sharing, mutual coordination and the like in an expected application scene in the sharing stage are considered, so that a decision maker can be supported to conduct weapon equipment system top-level planning and decision making, and a specific scheme comprising equipment research and development task allocation, research and development time, research and development quantity and equipment combination selection can be rapidly obtained.
In one embodiment, as shown in FIG. 2, the process of obtaining a feature vector matrix through a hierarchical multi-headed attention mechanism includes the steps of:
step 201, selecting a plurality of equipment from a set of equipment to be researched and developed to perform research and development work according to equipment development planning by taking a co-construction strategy as a drive;
Step 202, selecting a plurality of research and development units from the alternative research and development units to respectively bear different research and development tasks based on balanced meeting of different capacity requirements, and distributing equipment research and development types according to technical requirements of equipment research and development;
And 203, developing a spending budget according to the equipment cooperation requirement, taking the constraint of the increase of the equipment amplitude number on the capacity improvement amplitude as a basis, taking a sharing strategy as a drive, and designing a sharing co-building combination model based on development time, development spending and capacity redundancy.
Specifically, the complex equipment combination optimization problem of co-building shared dual-layer policy driving can be described as follows: when equipment development planning is carried out, a co-building strategy is used as a drive, a plurality of complex equipment is selected from a complex equipment set to be researched and developed for research and development, and a plurality of research and development units are selected from alternative research and development units to respectively bear different research and development tasks so as to balance and meet different capacity demands. According to the technical requirements required by the development of different equipment, the complex equipment is divided into several major categories, and more than one capability requirement can be met by different types of complex equipment. Different research and development units have different emphasis and strength, and the affordable armed equipment has different research and development types and different engineering amounts of research and development tasks. Meanwhile, constraints such as equipment cooperation, expense budget, different magnitude of capacity improvement and the like are considered, and the number of different magnitudes is increased, so that the research and development time is minimized as much as possible, and the expense and the capacity redundancy are developed. When the equipment combination is selected, the sharing strategy is used as a drive, and complex equipment is reasonably distributed according to the equipment configuration current situation and the mission tasks of different building units so as to maximally meet the capability requirements of the building units.
In one embodiment, as shown in fig. 3, the process of providing the feature vector matrix with a gaussian distribution includes:
Step 301, comprehensively considering and synchronously solving and optimizing a multi-stage equipment planning problem by adopting DNNB & BH algorithm by means of a probability priority decision mode and a tree search exploration mode;
Step 302, acquiring the self-association of the equipment combination scheme by adopting the internal structure and the learning mechanism of the deep neural network, and utilizing the learning neural network model to assist the solving process so as to reflect human factors into the solving process;
Step 303, accelerating the speed of approaching the optimal solution in the solution through the branch decision, trimming redundant branches through the delimitation decision, reducing the space-knowing scale, and obtaining the feasible solution of the shared co-building combination model.
Specifically, by modeling and describing the multi-stage problem, DNNB & BH algorithm (Deep Neural Network assisted Branch and Bound Heuristic Algorithm) is collectively referred to as a branch-and-bound heuristic aided by a neural network to solve synchronously during the optimization process. The DNNB & BH algorithm integrates a deep neural network model into the tree search process of the branch-and-bound method to decide which branch to choose next and which branches to prune when reaching a new node in the search process. The deep neural network model achieves the purpose of assisting in making decisions for selecting and pruning branches in the searching process through supervised learning of various historical data such as expert experience, decision maker preference and the like and planning schemes. The core idea of the method is to consider each feasible solution as a matrix, and change the matrix through a defined neighborhood structure so as to gradually approach the optimal solution. In the whole process, all possibilities are explored by utilizing tree search, the exploration sequence is determined by assisting with a deep neural network model, and the search tree is simplified. Every time the search process enters an unexplored node, the deep neural network model determines the next searched node by evaluating the child nodes obtained by several given change strategies and predicting which of the child nodes obtained by the change strategies has the highest probability of leading to the best solution. And simultaneously, the lower bounds of all the child nodes are predicted, and the nodes corresponding to the lower bounds which are higher than the objective function value of the current optimal solution are trimmed, so that the purposes of simplifying the search tree and shortening the search time are achieved.
Preferably, a branch-and-bound approach can be used to solve the optimization problem. Branch-and-bound methods are always performed around the search tree, starting from the root node, by systematically exploring the children of the root node and their successor nodes. The branching process is a process of adding child nodes to the tree. The delimitation is to check the lower bound of the sub-problem during branching, and if the sub-node cannot generate a solution that is better than the current optimal solution, then the branch is pruned. The entire search flow ends until all child nodes cannot produce a better solution. The solution to a given optimization problem can be understood as a leaf node in the tree.
For a given feasible solution S i, a new feasible solution S i+1 may be obtained by applying some variance to it, then this new solution is the neighborhood solution of the feasible solution S i, the variance may be referred to as a neighborhood variance V, and V acts on the neighborhood solutions obtained in S i to obtain a set of V (S i), then we may define the following neighborhood structure for the complex equipment combination optimization problem driven by the co-construction shared bilayer strategy:
Neighborhood 1: assigning a unit j originally not bearing the development of the equipment i to the development task of the equipment, wherein the set of neighborhood changes is V 1(Si;
Neighborhood 2: canceling the research and development task of the unit j originally bearing the research and development of the equipment i for the equipment, wherein the set of the neighborhood change is V 2(Si);
Neighborhood 3: exchanging two development units j 1,j2 for the development task of equipment i, the set of neighborhood changes being V 3(Si);
Neighborhood 4: assigning equipment i to a unit k that did not receive the equipment, the set of neighborhood changes being V 4(Si);
Neighborhood 5: the equipment that would have been assigned to the unit is cancelled, and the set of neighborhood changes is V 5(Si).
The field changes 1,2,3 are directed to the feasible solutions under the co-building strategy, the field changes 4,5 are directed to the feasible solutions under the sharing strategy, and meanwhile, the field changes 1,4 and the field changes 2,5 have a one-to-one correspondence respectively, which means that when the feasible solution S i under the co-building strategy is applied with the field change 1, the corresponding feasible solution S i ' under the sharing strategy must generate the field change 4. This ensures that the solution space is explored while being driven by the co-building shared bi-layer strategy.
In this embodiment we use a regression network model. Wherein the branch decision depth neural network model is used to predict the probability that the incoming solution may be close to the optimal solution, and the branch pruning depth neural network model is used to predict the optimal lower limit of the incoming solution. We define a "distance" and a "probability" respectively to illustrate the meaning of the probability that a solution may be close to the optimal solution.
Theorem 1 (distance): the current solution translates to the number of changes needed for the optimal solution through neighborhood changes.
Theorem 2 (probability): the current solution may be translated into the possibility of an optimal solution by neighborhood changes.
Accordingly, the shorter the distance between the current solution and the optimal solution, the higher the probability, and the greater the likelihood that the current solution can obtain the optimal solution through searching. The following describes how a deep neural network model that makes branch decisions functions in a branch-and-bound approach. When the searching progress reaches a certain node n, a change strategy is sequentially adopted for matrix transformation on the node, the node obtained after transformation is n k, and a solution matrix s k,sk corresponding to the node is used as input to be transmitted into a branch decision depth neural network model. An output is obtained through the branch decision model, and the output represents the probability value corresponding to the corresponding node n k. These outputs are then used to determine which branches of node n should be explored (e.g., first explore the branch where the node with the highest corresponding probability value is located). The presence of probability values not only provides a ranking of the child nodes, but may also represent the confidence that each child node is given by the branch decision model.
The bounding decision depth neural network model has a similar structure to the branch decision model. Both models use the same input (solution matrix s k corresponding to node n k), except that the output of the bounding decision model represents the predicted best lower bound for that branch, which means the size of the final best objective function value that would be obtained if the subsequent search were down that branch. Therefore, if the predicted optimal lower limit exceeds or equals the objective function value of the optimal solution that has been currently found, then the branch has no value to continue exploration. Since the bounding decision model fluctuates slightly during prediction, it can be multiplied by a weight between 0 and 1 to reduce its heuristic lower limit.
In one embodiment, as shown in fig. 4, the process of performing cluster analysis based on the distance metric algorithm includes:
Step 401, obtaining a neighborhood solution set under different change strategies according to a feasible solution given by the shared co-building combination model;
Step 402, inputting the solution matrix of the neighborhood solution set into respective branch decision depth neural network models, obtaining the probability of converting the neighborhood change into the optimal solution, and taking the result with the highest probability in the set obtained by the neighborhood change as a child node of the neighborhood structure branch;
Step 403, inputting the child node into a given decision depth neural network model again, determining whether to reserve the branch corresponding to the child node, and determining whether to repair the branch corresponding to the child node;
and 404, when the stopping requirement is met, selecting one of the sub-nodes in the search tree with highest corresponding probability under the co-construction strategy as the current solution under the co-construction strategy, and simultaneously selecting the corresponding sub-node in the search tree with the sharing strategy as the current solution under the sharing strategy.
Specifically, the DNNB & BH algorithm solving steps for the complex equipment combination optimization problem design driven by the co-construction shared double-layer strategy are as follows:
First, an initial solution S 0 under the co-building strategy and an initial optimal solution S 0 ' under the sharing strategy corresponding to S 0 are generated, and the initial solution S 0,S0 ' is used as a current solution S a,Sa ' to enter respective search trees.
Secondly, three kinds of neighborhood changes are sequentially applied to the current solution S a under the co-construction strategy, the set of neighborhood structures obtained through the neighborhood changes 1,2 and 3 are V 1(Sa),V2(Sa) and V 3(Sa), two kinds of neighborhood changes are applied to the current solution S a 'under the sharing strategy, and the neighborhood structures obtained through the neighborhood changes 4 and 5 are V 4(Sa'),V5(Sa').
Third, inputting all solution matrices in sets V 1(Sa),V2(Sa),V3(Sa) and V 4(Sa'),V5(Sa') into respective branch-decision depth neural network models, these data will propagate through the neural network and reach the output layer where we use the softmax activation function to obtain results between 0 and 1, which can be considered as probabilities. Then, the result with the highest probability in the set obtained by each neighborhood change is left to serve as a child node of the neighborhood structure branch, so that in a search tree under a co-building strategy, 3 child nodes are corresponding to each father node, two neighborhood structures are corresponding to each father node in a search tree under a sharing strategy, two child nodes are corresponding to each father node, according to the neighborhood corresponding relation mentioned in the above, the child node obtained by the neighborhood change 1 corresponds to the child node obtained by the neighborhood change 4, and the child node obtained by the neighborhood change 2 corresponds to the child node obtained by the neighborhood change 5.
Inputting the sub-node obtained by the search tree under the co-building strategy into the given decision depth neural network model again, comparing the output with the current found optimal objective function value, if the output is smaller than the current found optimal objective function value, retaining the branch corresponding to the sub-node, otherwise, trimming the branch corresponding to the sub-node, and simultaneously trimming the branch corresponding to the sub-node in the shared strategy search tree.
Fifthly, judging whether the stopping requirement is met, if not, selecting one of the sub-nodes in the search tree with highest corresponding probability under the co-construction strategy, taking the selected one as a current solution S a under the co-construction strategy, simultaneously selecting the corresponding sub-node in the search tree under the sharing strategy, taking the selected sub-node as a current solution S a' under the sharing strategy, turning to the step 2, if yes, stopping the algorithm, and outputting a result.
It should be understood that, although the steps in the above-described flowcharts are shown in order as indicated by the arrows, these steps are not necessarily performed in order as indicated by the arrows. The steps are not strictly limited in order unless explicitly stated in the present embodiment, and may be performed in other orders. Moreover, at least some of the steps in the flowcharts described above may include a plurality of sub-steps or stages that are not necessarily performed at the same time, but may be performed at different times, and the order of execution of the sub-steps or stages is not necessarily sequential, but may be performed alternately or alternately with at least a part of the sub-steps or stages of other steps or other steps.
In one embodiment, as shown in fig. 5, there is provided an equipment combination system based on co-building sharing, comprising:
The shared co-building module 501 is configured to construct a shared co-building combined model from a set of equipment to be developed according to technical requirements and types of the development equipment and according to the number of alternative development units and equipment units to be received;
The algorithm construction module 502 is configured to design DNNB & BH algorithm according to the deep neural network and for the shared co-construction combined model;
A model solving module 503, configured to perform overall association and calculation on the shared co-building combined model through the DNNB & BH algorithm, so as to obtain a feasible solution of the shared co-building combined model;
and the equipment combination module 504 is used for obtaining the specific schemes of equipment development task allocation, development time, development quantity and equipment combination selection according to the feasible solutions.
In one embodiment, as shown in fig. 5, the shared co-building module 501 includes an equipment planning unit 5011, the equipment planning unit 5011 being configured to:
According to equipment development planning, taking a co-building strategy as a drive, and selecting a plurality of pieces of equipment from equipment sets to be researched and developed to perform research and development work;
selecting a plurality of research and development units from the alternative research and development units to respectively bear different research and development tasks based on balanced meeting of different capacity requirements, and distributing research and development types of equipment according to technical requirements of research and development of the equipment;
According to the equipment cooperation requirement, the development expense budget, the constraint of the increase of the equipment amplitude quantity to the capacity improvement amplitude is taken as a basis, the sharing strategy is taken as a drive, and the shared co-building combination model based on the development time, the development expense and the capacity redundancy is designed.
In one embodiment, as shown in fig. 5, the model solving module 503 includes a neural network auxiliary unit 5031, and the neural network auxiliary unit 5031 is configured to:
Adopting DNNB & BH algorithm to comprehensively consider and synchronously solve and optimize the multi-stage equipment planning problem by means of a probability priority decision mode and a tree search exploration mode;
The inherent structure and the learning mechanism of the deep neural network are adopted, the self-relevance of the equipment combination scheme is obtained, the learning neural network model is utilized to assist the solving process, and the human factors are reflected in the solving process;
And accelerating the speed of approaching the optimal solution in the solution through the branch decision, trimming redundant branches through the delimitation decision, reducing the space-knowing scale, and obtaining the feasible solution of the shared co-building combined model.
In one embodiment, as shown in fig. 5, the model solving module 503 includes a branch-and-bound unit 5032, where the branch-and-bound unit 5032 is configured to:
obtaining a neighborhood solution set under different change strategies according to the feasible solutions given by the shared co-building combination model;
Inputting the solution matrix of the neighborhood solution set into respective branch decision depth neural network models, acquiring probability of converting the neighborhood change into the optimal solution, and taking the result with highest probability in the set obtained by the neighborhood change as a child node of a neighborhood structure branch;
Inputting the child node into a given decision depth neural network model again, determining whether to reserve the branch corresponding to the child node, and determining whether to repair the branch corresponding to the child node;
When the stopping requirement is met, selecting one of the sub-nodes in the search tree under the co-construction strategy, which has the highest corresponding probability, as the current solution under the co-construction strategy, and simultaneously selecting the corresponding sub-node in the search tree under the sharing strategy, and taking the current solution under the sharing strategy as the current solution.
For specific limitations on the co-construction share-based equipment combining system, reference may be made to the above limitation on the co-construction share-based equipment combining method, and no further description is given here. The above-described individual modules in the equipment combination system based on co-building sharing may be implemented in whole or in part by software, hardware, and combinations thereof. The above modules may be embedded in hardware or may be independent of a processor in the computer device, or may be stored in software in a memory in the computer device, so that the processor may call and execute operations corresponding to the above modules.
FIG. 6 illustrates an internal block diagram of a computer device in one embodiment. As shown in fig. 6, the computer device includes a processor, a memory, a network interface, an input device, and a display screen connected by a system bus. The memory includes a nonvolatile storage medium and an internal memory. The non-volatile storage medium of the computer device stores an operating system, and may also store a computer program that, when executed by a processor, may cause the processor to implement a method of equipment assembly based on co-built sharing. The internal memory may also store a computer program that, when executed by the processor, causes the processor to perform an equipment assembly method based on co-construction sharing. The display screen of the computer equipment can be a liquid crystal display screen or an electronic ink display screen, the input device of the computer equipment can be a touch layer covered on the display screen, can also be keys, a track ball or a touch pad arranged on the shell of the computer equipment, and can also be an external keyboard, a touch pad or a mouse and the like.
It will be appreciated by those skilled in the art that the structure shown in FIG. 6 is merely a block diagram of some of the structures associated with the present inventive arrangements and is not limiting of the computer device to which the present inventive arrangements may be applied, and that a particular computer device may include more or fewer components than shown, or may combine some of the components, or have a different arrangement of components.
In one embodiment, a computer device is provided comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, the processor implementing the steps of when executing the computer program:
From a set of equipment to be developed, constructing a shared co-building combined model according to technical requirements and types of the development equipment and according to the number of alternative development units and equipment units to be received;
aiming at the shared co-building combination model, designing DNNB & BH algorithm according to the deep neural network;
Carrying out overall association and calculation on the shared co-building combined model through the DNNB & BH algorithm to obtain a feasible solution of the shared co-building combined model;
and according to the feasible solution, obtaining the specific schemes of equipment development task allocation, development time, development quantity and equipment combination selection.
In one embodiment, the processor when executing the computer program further performs the steps of:
According to equipment development planning, taking a co-building strategy as a drive, and selecting a plurality of pieces of equipment from equipment sets to be researched and developed to perform research and development work;
selecting a plurality of research and development units from the alternative research and development units to respectively bear different research and development tasks based on balanced meeting of different capacity requirements, and distributing research and development types of equipment according to technical requirements of research and development of the equipment;
According to the equipment cooperation requirement, the development expense budget, the constraint of the increase of the equipment amplitude quantity to the capacity improvement amplitude is taken as a basis, the sharing strategy is taken as a drive, and the shared co-building combination model based on the development time, the development expense and the capacity redundancy is designed.
In one embodiment, the processor when executing the computer program further performs the steps of:
Adopting DNNB & BH algorithm to comprehensively consider and synchronously solve and optimize the multi-stage equipment planning problem by means of a probability priority decision mode and a tree search exploration mode;
The inherent structure and the learning mechanism of the deep neural network are adopted, the self-relevance of the equipment combination scheme is obtained, the learning neural network model is utilized to assist the solving process, and the human factors are reflected in the solving process;
And accelerating the speed of approaching the optimal solution in the solution through the branch decision, trimming redundant branches through the delimitation decision, reducing the space-knowing scale, and obtaining the feasible solution of the shared co-building combined model.
In one embodiment, the processor when executing the computer program further performs the steps of:
Obtaining a neighborhood solution set under different change strategies according to the feasible solutions given by the shared co-building combination model;
Inputting the solution matrix of the neighborhood solution set into respective branch decision depth neural network models, acquiring probability of converting the neighborhood change into the optimal solution, and taking the result with highest probability in the set obtained by the neighborhood change as a child node of a neighborhood structure branch;
Inputting the child node into a given decision depth neural network model again, determining whether to reserve the branch corresponding to the child node, and determining whether to repair the branch corresponding to the child node;
When the stopping requirement is met, selecting one of the sub-nodes in the search tree under the co-construction strategy, which has the highest corresponding probability, as the current solution under the co-construction strategy, and simultaneously selecting the corresponding sub-node in the search tree under the sharing strategy, and taking the current solution under the sharing strategy as the current solution.
In one embodiment, a computer device is provided comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, the processor implementing the steps of when executing the computer program:
From a set of equipment to be developed, constructing a shared co-building combined model according to technical requirements and types of the development equipment and according to the number of alternative development units and equipment units to be received;
aiming at the shared co-building combination model, designing DNNB & BH algorithm according to the deep neural network;
Carrying out overall association and calculation on the shared co-building combined model through the DNNB & BH algorithm to obtain a feasible solution of the shared co-building combined model;
and according to the feasible solution, obtaining the specific schemes of equipment development task allocation, development time, development quantity and equipment combination selection.
In one embodiment, the processor when executing the computer program further performs the steps of:
According to equipment development planning, taking a co-building strategy as a drive, and selecting a plurality of pieces of equipment from equipment sets to be researched and developed to perform research and development work;
selecting a plurality of research and development units from the alternative research and development units to respectively bear different research and development tasks based on balanced meeting of different capacity requirements, and distributing research and development types of equipment according to technical requirements of research and development of the equipment;
According to the equipment cooperation requirement, the development expense budget, the constraint of the increase of the equipment amplitude quantity to the capacity improvement amplitude is taken as a basis, the sharing strategy is taken as a drive, and the shared co-building combination model based on the development time, the development expense and the capacity redundancy is designed.
In one embodiment, the processor when executing the computer program further performs the steps of:
Adopting DNNB & BH algorithm to comprehensively consider and synchronously solve and optimize the multi-stage equipment planning problem by means of a probability priority decision mode and a tree search exploration mode;
The inherent structure and the learning mechanism of the deep neural network are adopted, the self-relevance of the equipment combination scheme is obtained, the learning neural network model is utilized to assist the solving process, and the human factors are reflected in the solving process;
And accelerating the speed of approaching the optimal solution in the solution through the branch decision, trimming redundant branches through the delimitation decision, reducing the space-knowing scale, and obtaining the feasible solution of the shared co-building combined model.
In one embodiment, the processor when executing the computer program further performs the steps of:
Obtaining a neighborhood solution set under different change strategies according to the feasible solutions given by the shared co-building combination model;
Inputting the solution matrix of the neighborhood solution set into respective branch decision depth neural network models, acquiring probability of converting the neighborhood change into the optimal solution, and taking the result with highest probability in the set obtained by the neighborhood change as a child node of a neighborhood structure branch;
Inputting the child node into a given decision depth neural network model again, determining whether to reserve the branch corresponding to the child node, and determining whether to repair the branch corresponding to the child node;
When the stopping requirement is met, selecting one of the sub-nodes in the search tree under the co-construction strategy, which has the highest corresponding probability, as the current solution under the co-construction strategy, and simultaneously selecting the corresponding sub-node in the search tree under the sharing strategy, and taking the current solution under the sharing strategy as the current solution.
Those skilled in the art will appreciate that implementing all or part of the above described methods may be accomplished by way of a computer program stored on a non-transitory computer readable storage medium, which when executed, may comprise the steps of the embodiments of the methods described above.
The technical features of the above embodiments may be arbitrarily combined, and all possible combinations of the technical features in the above embodiments are not described for brevity of description, however, as long as there is no contradiction between the combinations of the technical features, they should be considered as the scope of the description.
The above examples illustrate only a few embodiments of the application, which are described in detail and are not to be construed as limiting the scope of the application. It should be noted that it will be apparent to those skilled in the art that several variations and modifications can be made without departing from the spirit of the application, which are all within the scope of the application. Accordingly, the scope of protection of the present application is to be determined by the appended claims.

Claims (4)

1. A method of equipment assembly based on co-construction sharing, the method comprising:
From a set of equipment to be developed, constructing a shared co-building combined model according to technical requirements and types of the development equipment and according to the number of alternative development units and equipment units to be received;
aiming at the shared co-building combination model, designing DNNB & BH algorithm according to the deep neural network;
Carrying out overall association and calculation on the shared co-building combined model through the DNNB & BH algorithm to obtain a feasible solution of the shared co-building combined model;
Acquiring specific schemes of equipment research and development task allocation, research and development time, research and development quantity and equipment combination selection according to the feasible solutions;
The building of the shared co-building combination model from the to-be-developed equipment set according to the technical requirements and types of the to-be-developed equipment and according to the number of the alternative development units and the to-be-received equipment units comprises the following steps:
According to equipment development planning, taking a co-building strategy as a drive, and selecting a plurality of pieces of equipment from equipment sets to be researched and developed to perform research and development work;
selecting a plurality of research and development units from the alternative research and development units to respectively bear different research and development tasks based on balanced meeting of different capacity requirements, and distributing research and development types of equipment according to technical requirements of research and development of the equipment;
According to the equipment cooperation requirement, developing a spending budget, taking the constraint of the increase of the number of the equipment amplitudes on the capacity lifting amplitude as a basis, taking a sharing strategy as a drive, and designing a sharing co-building combination model based on development time, developing spending and capacity redundancy;
And performing overall association and calculation on the shared co-building combination model through the DNNB & BH algorithm to obtain a feasible solution of the shared co-building combination model, wherein the method comprises the following steps:
Adopting DNNB & BH algorithm to comprehensively consider and synchronously solve and optimize the multi-stage equipment planning problem by means of a probability priority decision mode and a tree search exploration mode;
The inherent structure and the learning mechanism of the deep neural network are adopted, the self-relevance of the equipment combination scheme is obtained, the learning neural network model is utilized to assist the solving process, and the human factors are reflected in the solving process;
Accelerating the speed of approaching the optimal solution in the solving process through the branch decision, trimming redundant branches through the delimitation decision, reducing the space knowing scale and obtaining the feasible solution of the shared co-building combined model;
And performing overall association and calculation on the shared co-building combination model through the DNNB & BH algorithm to obtain a feasible solution of the shared co-building combination model, wherein the method comprises the following steps:
obtaining a neighborhood solution set under different change strategies according to the feasible solutions given by the shared co-building combination model;
Inputting the solution matrix of the neighborhood solution set into respective branch decision depth neural network models, acquiring probability of converting the neighborhood change into the optimal solution, and taking the result with highest probability in the set obtained by the neighborhood change as a child node of a neighborhood structure branch;
Inputting the child node into a given decision depth neural network model again, determining whether to reserve the branch corresponding to the child node, and determining whether to repair the branch corresponding to the child node;
When the stopping requirement is met, selecting one of the sub-nodes in the search tree under the co-building strategy, which has the highest corresponding probability, as the current solution under the co-building strategy, and simultaneously selecting the corresponding sub-node in the search tree under the sharing strategy, and taking the current solution under the sharing strategy as the current solution;
the solution step of the complex equipment combination optimization problem design driven by the co-construction shared double-layer strategy through DNNB & BH algorithm comprises the following steps:
Firstly, generating an initial solution S 0 under a co-building strategy and an initial optimal solution S 0 ' under a sharing strategy corresponding to S 0, wherein the initial solution S 0,S0 ' is used as a current solution S a,Sa ' to enter respective search trees;
Secondly, three types of neighborhood changes are sequentially applied to a current solution S a under the co-construction strategy, the set of neighborhood structures obtained through the neighborhood changes 1,2 and 3 are V 1 (Sa), V2 (Sa) and V 3 (Sa) respectively, two types of neighborhood changes are applied to a current solution S a 'under the sharing strategy respectively, and the neighborhood structures obtained through the neighborhood changes 4 and 5 are V 4 (Sa'),V5 (Sa') respectively;
third, inputting all solution matrices in sets V 1 (Sa), V2 (Sa ),V3 (Sa) and V 4 (Sa'),V5 (Sa') into respective branch-decision depth neural network models, these data will propagate through the neural network and reach the output layer where the softmax activation function is used to obtain results between 0 and 1, which are considered as probabilities; then, the result with the highest probability in the set obtained by each neighborhood change is left to serve as a child node of the neighborhood structure branch, so that 3 child nodes are corresponding to each father node in a search tree under a co-building strategy, two neighborhood structures are corresponding to each father node in a search tree under a sharing strategy, two child nodes are corresponding to each father node, the child node obtained by the neighborhood change 1 corresponds to the child node obtained by the neighborhood change 4 according to the neighborhood corresponding relation mentioned in the above, and the child node obtained by the neighborhood change 2 corresponds to the child node obtained by the neighborhood change 5;
Inputting the sub-node obtained by the search tree under the co-building strategy into the given decision depth neural network model again, comparing the output with the current found optimal objective function value, if the output is smaller than the current found optimal objective function value, reserving the branch corresponding to the sub-node, otherwise, trimming the branch corresponding to the sub-node, and simultaneously trimming the branch corresponding to the sub-node in the shared strategy search tree;
Fifthly, judging whether the stopping requirement is met, if not, selecting one of the sub-nodes in the search tree with highest corresponding probability under the co-construction strategy, taking the selected one as a current solution S a under the co-construction strategy, simultaneously selecting the corresponding sub-node in the search tree under the sharing strategy, taking the selected sub-node as a current solution S a' under the sharing strategy, turning to the step 2, if yes, stopping the algorithm, and outputting a result.
2. An equipment combining system based on co-construction sharing, comprising:
The shared co-building module is used for building a shared co-building combined model from a to-be-researched and developed equipment set according to the technical requirements and types of research and development equipment and according to the number of alternative research and development units and to-be-received equipment units;
The algorithm construction module is used for aiming at the shared co-construction combined model and designing DNNB & BH algorithm according to the deep neural network;
The model solving module is used for carrying out overall association and calculation on the shared co-built combined model through the DNNB & BH algorithm to obtain a feasible solution of the shared co-built combined model;
The equipment combination module is used for obtaining specific schemes of equipment research and development task allocation, research and development time, research and development quantity and equipment combination selection according to the feasible solution;
the shared co-building module includes an equipment planning unit for:
According to equipment development planning, taking a co-building strategy as a drive, and selecting a plurality of pieces of equipment from equipment sets to be researched and developed to perform research and development work;
selecting a plurality of research and development units from the alternative research and development units to respectively bear different research and development tasks based on balanced meeting of different capacity requirements, and distributing research and development types of equipment according to technical requirements of research and development of the equipment;
According to the equipment cooperation requirement, developing a spending budget, taking the constraint of the increase of the number of the equipment amplitudes on the capacity lifting amplitude as a basis, taking a sharing strategy as a drive, and designing a sharing co-building combination model based on development time, developing spending and capacity redundancy;
the model solving module includes a neural network assistance unit for:
Adopting DNNB & BH algorithm to comprehensively consider and synchronously solve and optimize the multi-stage equipment planning problem by means of a probability priority decision mode and a tree search exploration mode;
The inherent structure and the learning mechanism of the deep neural network are adopted, the self-relevance of the equipment combination scheme is obtained, the learning neural network model is utilized to assist the solving process, and the human factors are reflected in the solving process;
Accelerating the speed of approaching the optimal solution in the solving process through the branch decision, trimming redundant branches through the delimitation decision, reducing the space knowing scale and obtaining the feasible solution of the shared co-building combined model;
the model solving module comprises a branch-and-bound unit for:
obtaining a neighborhood solution set under different change strategies according to the feasible solutions given by the shared co-building combination model;
Inputting the solution matrix of the neighborhood solution set into respective branch decision depth neural network models, acquiring probability of converting the neighborhood change into the optimal solution, and taking the result with highest probability in the set obtained by the neighborhood change as a child node of a neighborhood structure branch;
Inputting the child node into a given decision depth neural network model again, determining whether to reserve the branch corresponding to the child node, and determining whether to repair the branch corresponding to the child node;
When the stopping requirement is met, selecting one of the sub-nodes in the search tree under the co-building strategy, which has the highest corresponding probability, as the current solution under the co-building strategy, and simultaneously selecting the corresponding sub-node in the search tree under the sharing strategy, and taking the current solution under the sharing strategy as the current solution;
The method adopts DNNB & BH algorithm to solve the complex equipment combination optimization problem design driven by co-construction sharing double-layer strategy, and comprises the following steps:
Firstly, generating an initial solution S 0 under a co-building strategy and an initial optimal solution S 0 ' under a sharing strategy corresponding to S 0, wherein the initial solution S 0,S0 ' is used as a current solution S a,Sa ' to enter respective search trees;
Secondly, three types of neighborhood changes are sequentially applied to a current solution S a under the co-construction strategy, the set of neighborhood structures obtained through the neighborhood changes 1,2 and 3 are V 1 (Sa), V2 (Sa) and V 3 (Sa) respectively, two types of neighborhood changes are applied to a current solution S a 'under the sharing strategy respectively, and the neighborhood structures obtained through the neighborhood changes 4 and 5 are V 4 (Sa'),V5 (Sa') respectively;
third, inputting all solution matrices in sets V 1 (Sa), V2 (Sa ),V3 (Sa) and V 4 (Sa'),V5 (Sa') into respective branch-decision depth neural network models, these data will propagate through the neural network and reach the output layer where the softmax activation function is used to obtain results between 0 and 1, which are considered as probabilities; then, the result with the highest probability in the set obtained by each neighborhood change is left to serve as a child node of the neighborhood structure branch, so that 3 child nodes are corresponding to each father node in a search tree under a co-building strategy, two neighborhood structures are corresponding to each father node in a search tree under a sharing strategy, two child nodes are corresponding to each father node, the child node obtained by the neighborhood change 1 corresponds to the child node obtained by the neighborhood change 4 according to the neighborhood corresponding relation mentioned in the above, and the child node obtained by the neighborhood change 2 corresponds to the child node obtained by the neighborhood change 5;
Inputting the sub-node obtained by the search tree under the co-building strategy into the given decision depth neural network model again, comparing the output with the current found optimal objective function value, if the output is smaller than the current found optimal objective function value, reserving the branch corresponding to the sub-node, otherwise, trimming the branch corresponding to the sub-node, and simultaneously trimming the branch corresponding to the sub-node in the shared strategy search tree;
Fifthly, judging whether the stopping requirement is met, if not, selecting one of the sub-nodes in the search tree with highest corresponding probability under the co-construction strategy, taking the selected one as a current solution S a under the co-construction strategy, simultaneously selecting the corresponding sub-node in the search tree under the sharing strategy, taking the selected sub-node as a current solution S a' under the sharing strategy, turning to the step 2, if yes, stopping the algorithm, and outputting a result.
3. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the steps of the method of claim 1 when executing the computer program.
4. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the method of claim 1.
CN202111315568.9A 2021-11-08 2021-11-08 Equipment combination method, system, equipment and storage medium based on co-construction sharing Active CN114036837B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111315568.9A CN114036837B (en) 2021-11-08 2021-11-08 Equipment combination method, system, equipment and storage medium based on co-construction sharing

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111315568.9A CN114036837B (en) 2021-11-08 2021-11-08 Equipment combination method, system, equipment and storage medium based on co-construction sharing

Publications (2)

Publication Number Publication Date
CN114036837A CN114036837A (en) 2022-02-11
CN114036837B true CN114036837B (en) 2024-06-04

Family

ID=80143477

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111315568.9A Active CN114036837B (en) 2021-11-08 2021-11-08 Equipment combination method, system, equipment and storage medium based on co-construction sharing

Country Status (1)

Country Link
CN (1) CN114036837B (en)

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108256671A (en) * 2017-12-26 2018-07-06 佛山科学技术学院 A kind of more resources of multitask based on learning-oriented genetic algorithm roll distribution method
CN108364138A (en) * 2018-03-15 2018-08-03 中国人民解放军国防科技大学 Weapon equipment development planning modeling and solving method based on countermeasure visual angle
CN110989343A (en) * 2019-11-25 2020-04-10 中国人民解放军国防科技大学 Multi-stage equipment combination planning method based on reinforcement learning
CN111476490A (en) * 2020-04-08 2020-07-31 郑州天迈科技股份有限公司 Regional multi-line vehicle scheduling algorithm shared by resource pool
CN112380434A (en) * 2020-11-16 2021-02-19 吉林大学 Interpretable recommendation system method fusing heterogeneous information network
CN112884289A (en) * 2021-01-25 2021-06-01 中国人民解放军国防科技大学 Weapon and equipment combination selection method and system based on system contribution rate
CN112926943A (en) * 2021-03-11 2021-06-08 中国人民解放军国防科技大学 Automatic scheduling method, system, computer equipment and storage medium

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108256671A (en) * 2017-12-26 2018-07-06 佛山科学技术学院 A kind of more resources of multitask based on learning-oriented genetic algorithm roll distribution method
CN108364138A (en) * 2018-03-15 2018-08-03 中国人民解放军国防科技大学 Weapon equipment development planning modeling and solving method based on countermeasure visual angle
CN110989343A (en) * 2019-11-25 2020-04-10 中国人民解放军国防科技大学 Multi-stage equipment combination planning method based on reinforcement learning
CN111476490A (en) * 2020-04-08 2020-07-31 郑州天迈科技股份有限公司 Regional multi-line vehicle scheduling algorithm shared by resource pool
CN112380434A (en) * 2020-11-16 2021-02-19 吉林大学 Interpretable recommendation system method fusing heterogeneous information network
CN112884289A (en) * 2021-01-25 2021-06-01 中国人民解放军国防科技大学 Weapon and equipment combination selection method and system based on system contribution rate
CN112926943A (en) * 2021-03-11 2021-06-08 中国人民解放军国防科技大学 Automatic scheduling method, system, computer equipment and storage medium

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
武器装备体系规划备选方案生成方法;郭栋;张迎新;韩高飞;周洋臣;;指挥控制与仿真;20201231(第05期);全文 *
面向武器装备体系组合规划的集成决策优化框架;周宇;杨克巍;姜江;谭跃进;;国防科技大学学报;20130628(第03期);全文 *
面向能力需求的武器装备体系组合规划模型与算法;周宇;谭跃进;姜江;杨克巍;赵青松;;系统工程理论与实践;20130315(第03期);全文 *

Also Published As

Publication number Publication date
CN114036837A (en) 2022-02-11

Similar Documents

Publication Publication Date Title
CN114915630B (en) Task allocation method, network training method and device based on Internet of Things equipment
CN103646008A (en) Web service combination method
Ekel et al. Multicriteria analysis in decision making under information uncertainty
CN110989343B (en) Multi-stage equipment combination planning method based on reinforcement learning
CN110297490A (en) Heterogeneous module robot via Self-reconfiguration planing method based on nitrification enhancement
CN110688754A (en) Combat system architecture modeling and optimal search method
CN108106624A (en) A kind of more people's Dispatch by appointment paths planning methods and relevant apparatus
CN115756478A (en) Method for automatically fusing operators of calculation graph and related product
Gunawan et al. ADOPT: Combining parameter tuning and Adaptive Operator Ordering for solving a class of Orienteering Problems
CN116992151A (en) Online course recommendation method based on double-tower graph convolution neural network
CN114036837B (en) Equipment combination method, system, equipment and storage medium based on co-construction sharing
EP0517036A2 (en) A development supporting system for fuzzy inference device and the scheme by which this system operates
Qu et al. A comprehensive review of machine learning in multi-objective optimization
Hongyun et al. Multi-goal path planning algorithm for mobile robots in grid space
Nguyen et al. Tradeoff different construction project goals in using a novel multi-objective sea horse algorithm
Sharma et al. Power law-based local search in artificial bee colony
CN116341634A (en) Training method and device for neural structure search model and electronic equipment
CN114707643A (en) Model segmentation method and related equipment thereof
Saha et al. How Does User Behavior Evolve During Exploratory Visual Analysis?
CN110675088B (en) Efficient division method for complex tasks of digital twin system
Eisenstadt et al. Autocompletion of Floor Plans for the Early Design Phase in Architecture: Foundations, Existing Methods, and Research Outlook
CN112190945A (en) Game data processing method and device, computer readable storage medium and electronic equipment
CN117951282B (en) Multi-information-fused graph neural network session recommendation method
Li et al. Discrete artificial bee colony algorithm with fixed neighborhood search for traveling salesman problem
CN113240189B (en) Reputation value-based dynamic vehicle task and calculation force matching method

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
GR01 Patent grant
GR01 Patent grant