Research On Weapon Equipment Portfolio Selection Method Based On System Of Systems Contribution Rate Assessment

Weapon and equipment development planning is a complex system engineering that determines the future development direction,scale structure,and capability level of various types of weapons and equipment.It relates to national security and future military struggles.It has important military significance and research value.The department-centered "chimney-style" weaponry development planning model and fragmented decision-making process have emerged many "similar/duplicate capabilities" and "difficult to interact/integrate" weaponry development projects,which not only caused a waste of resources for national defense investment,And it greatly hinders the process of "joint capability aggregation" of scattered weapons and equipment at the system level.From the perspective of the system of system,this paper takes the combined development plan of weapons and equipment as the background,and under the demand of actual military problems,researches on the method of weapon and equipment portfolio based on the system contribution rate.Combining complex systems and complex networks,portfolio analysis,multi-objective optimization,machine learning,and other theories and methods,“considering the condition of single-stage and singlescenario,multi-stage,multi-scenarios,and multi-objective,the research of weapon equipment portfolio for typical scenarios methods,the weapon and equipment portfolio considering dynamic evolution,the weapon and equipment portfolio considering the scene correlation,and the weapon and equipment portfolio method considering the development risk,etc.are carried out.The research work and innovations of the thesis include:(1)Proposed a modeling method of weapon equipment portfolio method based on the system contribution rateFully consider the relationship between the elements of the system and the heterogeneity of element types,and carry out networked modeling and description of the weapon equipment system based on the idea of a functional chain.Taking equipment portfolio as the object of contribution rate evaluation,analyzing the contribution rate of weapon equipment portfolio under the system background,and proposing a weapon equipment portfolio model based on the system contribution rate.The model fully considers the interrelationship between equipment,and under the condition of capability requirements and funding constraints,the equipment portfolio is chosen to maximize the combat capability of the entire weapon equipment system.(2)Proposed a method for weapons and equipment portfolio method considering dynamic evolutionDevelopment planning in reality often requires multiple stages of continuous development.In particular,the development plan for equipping this kind of complex product will generally be developed in multiple stages in a cycle of several years,and the budgetary funds will be allocated annually.Therefore,it is required that the weapon equipment development plan and the selection of the equipment to be developed should also be based on the multi-stage dynamic evolution characteristics and capability demand changes to make multi-stage decisions.In this paper,considering the future capacity requirements and the uncertainty of funding inputs at different stages,a multistage portfolio method for equipment based on the Dueling Deep Q-Network(Dueling DQN)deep reinforcement learning algorithm is proposed.This method can well deal with the inherent uncertainty faced by multi-stage dynamic evolution,making the weapon equipment development plan more reasonable.(3)A method for weapons and equipment portfolio method considering scene correlation is proposedIn reality,there are quite a lot of equipment or equipment combinations to be used in a variety of scenarios to serve diverse mission directions.In order to maximize the need for different missions in multiple scenarios,it is necessary to develop a combination of weapons and equipment for multiple scenarios.This paper presents a method for weapons and equipment portfolio that considers the correlation of scenes.For multiple possible scenarios in the future,considering the correlation between the scenes,the planning model of the equipment portfolio method based on the importance of the scene is constructed.First,calculate the scene relevance based on multidimensional information fusion;then,calculate the importance of each scene based on the Page Rank method,and then calculate the contribution rate of each equipment portfolio in each scene to obtain the contribution rate matrix.The combination of the two results in each equipment portfolio in all Contribution rates of the comprehensive system under the scenario;finally,the equipment combination plan is selected according to the contribution rate of the comprehensive system.(4)Proposed a multi-target selection method for weapon and equipment portfolio considering development risksFully consider the risks of cost overruns and development failure in the development of weapon equipment portfolios,expand the contribution rate research in multiple dimensions,and propose a multi-objective selection method for weapon equipment portfolios considering development risks.This method takes the contribution rate and development risk of the weapon equipment portfolio system as dual objectives to optimize and takes budget cost and capacity requirements as constraints.It uses an improved NSGA-II(Non-dominated Sorting Genetic Algorithm-II)hybrid intelligent optimization algorithm NSDE(Non-dominated Sorting Genetic Algorithm-II).-dominated Sorting Differential Evolution)to solve the multi-objective optimization problem and obtain a set of Pareto optimal solutions.Then,a refined selection strategy of non-dominated equipment portfolio based on non-dominated solution set association rule mining is proposed,and the Pareto optimal solution is comprehensively weighed to obtain the optimal equipment portfolio plan,which is compared with other intelligent optimization algorithms,which proves the advantages of the proposed method.The obtained optimal equipment portfolio plan not only increases the contribution rate of the system but also effectively avoids its development risks.