Research Of Belief-Rule-based Reasoning Technology For Wartime Equipment Maintenance

Several high-tech local wars in the past ten years have shown that the war is not only a contest of combat effectiveness between belligerents,but also a confrontation of equipment support.The quality of wartime equipment support will affect the process and result of war.Wartime equipment maintenance support is an important factor constituting combat effectiveness,which will play a more and more prominent role in future wars.Therefore,how to establish a wartime equipment support decision system that meets the needs of future war has become a research hotspot in the field of equipment support in various countries.Aiming at the issue of equipment maintenance support in wartime,this paper first establishes an appropriate set of indicators describing equipment maintenance support,and divides maintenance support tasks according to the components of equipment and the degree of equipment damage.Then,based on expert experience and equipment test data extraction rules,the corresponding belief rule base is further established,and optimize the parameters and structure of the belief rule base system,improve the speed and accuracy of the system’s reasoning,and build an equipment support decision-making model and optimization algorithm that conform to the actual battlefield.The main research innovations are reflected in the following aspects:1.Aiming at the shortcomings of the existing belief rule base parameter optimization models,this paper proposes a belief rule-base reasoning method based on improved second-order oscillatory particle swarm optimization.Based on the particle swarm optimization parameter training method,this method improves the particle velocity evolution strategy,introduces the second-order oscillation operator and adaptive random inertia weight,which effectively overcomes the problems of premature convergence and falling into local optimal solution of the algorithm,and improves the accuracy of the belief rule-base reasoning method.2.Most of the existing belief rule-base are constructed by traversal combination.The disadvantage is that as the number of premise attributes increases,the number of belief rules will inevitably increase exponentially,which will seriously affect the speed of reasoning.In wartime,equipment support decision-making requires rapid response,so in order to increase the speed of reasoning,it is necessary to reduce the rules of the belief rule-base.This paper constructs a belief rule library by linear combination,sets the number of rules equal to evaluation levels and improves the calculation formula of activation weight,which effectively overcomes the "combination explosion" problem of the number of rules and the "zero activation" problem of rules.3.This paper studies the typical equipment health status evaluation problem,and proposes a method for constructing equipment performance degradation indicators based on hidden Markov chains.And combined with the inference method of the belief rule-base to evaluate the health status of the equipment and realize the prediction of the remaining life.First,the hidden Markov model is used to obtain the equipment performance index degradation value of the multi-sensor fusion;then,the built-in degradation index is used to train the belief rule-base to identify the health status of the system,and then to predict its failure and remaining life.This provides new ideas and methods for equipment maintenance support modeling.4.In order to solve the problem of multiple sources and uncertainty of data in a complex battlefield environment,quickly and accurately assess the damage level of equipment.This paper proposes a method of evaluating equipment battle damage based on belief rule-base and evidence reasoning.In view of the increase in the number of battle damage evaluation levels,the accuracy of the evaluation will drop sharply.Based on the existing reasoning algorithm of the belief rule-base,this paper introduces a two-choice public warehouse decision model to construct multiple two-category belief rule-base to solve the same classification problem,which effectively improves the assessment accuracy of equipment battle damage levels,thereby laying the foundation for rapid and effective battlefield repair.At the end of this paper,we combine these methods to build a wartime equipment maintenance support prototype system,and simulation experiments are carried out.Simulation experiments show that the model and method proposed in this paper can effectively evaluate the health status of typical equipment systems and quickly assess equipment damage levels during wartime,and has been applied in the weapon equipment pre research project "Research on performance evaluation technology of XXX typical equipment for combat mission",so as to provide support for battlefield commanders to make equipment support decisions.