Research Of Equipment Performance Evaluation Technology Based On Belief Rule-base

Under the development trend of modern warfare,equipments have been integrated with various high-tech and become more and more complicated.Moreover,the new methods of warfare require the integration of various information equipment,and maintain the interconnection of equipment and combat systems,and put forward higher requirements for the equipment support for combat missions.In this context,this paper has carried out research on the evaluation method of equipment performance.Based on the background of equipment support for combat missions,this paper proposes and constructs an equipment performance evaluation system based on the belief rule-base to evaluate the performance of equipment.The system starts from the hierarchical index relationship of equipment,and builds a hierarchical belief rule-base system to evaluate equipment performance.Firstly,the rule extraction is based on the rough set;then,according to the integrity of the input information,the traditional belief rule-base reasoning method or the belief rule-base reasoning method of the incomplete input information proposed in this paper is determined;finally,based on the large amount of data,the equipment performance evaluation system is optimized,which is divided into belief rule-base parameter optimization based on adaptive multi-population genetic algorithm and belief rule-base structure optimization based on statistical utility.The paper studies the equipment performance evaluation problem,the construction process of the belief rule-base system and the problems encountered in it.The main research innovations are reflected in the following aspects:1.For the problem that the previous belief rule extraction relies solely on expert experience,a method based on rough set is proposed.The method is based on data,and the data is divided into equivalence classes according to the reference value of its attributes,and then rules is extracted according to the support degree,and finally optimize the number and spacing of attribute reference values by the degree of the rules,so that the extracted rules are more suitable for the actual performance evaluation of equipment.2.According to the hierarchical index relationship of equipment performance,a hierarchical belief rule-base system is constructed,and the evaluation result of the previous layer of belief rule-base system is processed,and it is used as the predecessor of the next layer of belief rule-base,which has certain innovation,so that the problem of inconsistent input and output forms of different levels of belief rule-base is solved.3.In the process of equipment performance evaluation,often due to human error,difficulty in key parameters acquisition,etc.,the input information is incomplete,and thus the reasoning cannot be completed normally.Based on this,this paper proposes a belief rule-base reasoning method based on distribution.This method supplements themissing attributes according to the distribution of data,forms multiple inputs,uses the belief rule-base reasoning method to reason,and finally uses ER algorithm to fuse all the inference results into a comprehensive inference result.This method can overcome the problem of reasoning while minimizing errors.4.In order to overcome difficulty of optimization due to the large difference between the different belief rule-bases in this paper,this paper proposes an adaptive multi-group genetic algorithm to optimize the belief rule-base.This method sets up multiple sub-populations to improve the global optimization ability of the algorithm,and there is a migration strategy between sub-populations.In the specific evolution process,adaptive crossover operators and adaptive mutation operators are added to speed up the evolution,thus improving the global optimization ability and local optimization ability of genetic algorithm.At the end of this paper,we combine these methods to build a performance evaluation prototype system for equipment support,and the simulation data is generated by the analytic hierarchy process.Using these simulation data,this paper verifies the effectiveness of the previous set of methods.Through experiments,it is found that the process of rule extraction,rule reasoning and rule optimization proposed in this paper is suitable for equipment performance evaluation,which can greatly improve the accuracy of equipment performance evaluation,and thus provide support for equipment support for combat missions.