Mission Reliability Modeling And Evaluation Technology For UAV Swarm System

With the advent of intelligence era,new weapons such as unmanned aerial vehicles have gradually become the protagonists of military confrontation.The ever-changing battlefield of the future makes the survivability of a single UAV limited.However,the UAV swarm can expand the capability of a single UAV and enhance the overall operational effectiveness.It is important in engineering application to evaluate the mission reliability of UAV swarm system.In this paper,based on the multi-phases presented in the task process of UAV swarm system,the mission reliability modeling and evaluation for such phased-mission systems(PMS)is studied.The main contents of this paper are as follows.(1)The development of PMS models based on binary decision diagram(BDD)for UAV swarm systemThe existing PMS studies mainly focus on a single platform.However,in the background of UAV swarm operation,it is necessary to consider the whole UAV swarm system.Therefore,based on the reported single platform PMS modeling method,we regarded the UAV swarm as a k-out-of-n system and developed the PMS model for the swarm system as well as the evaluation method.When modeling a single platform,the method of generating BDD based on minimal path set is adopted.The generated BDD can give the success status of the system directly,without combining the failure of current phase with the success of previous phase,which eliminates the process of getting logic BDDs from the first phase to the(n-1)th phase.The reliability prediction result of this method is shown to be consistent with the existing FT-based PMS method.(2)PMS modeling of swarm system with multi-state componentBased on the PMS modeling in(1),the multi-state performance of the component is considered.We regard the UAV swarm as a multi-state phased-mission system(MS-PMS)and study the multi-state PMS reliability modeling method for UAV swarm system.We divided the states of UAV component,and then gave the method to evaluate the mission reliability of the UAV swarm system under polymorphism based on the multistate multivalued decision diagram(MMDD).(3)Ordering heuristics for reliability evaluation method based on path vector using MMDDThe size of MMDD generated based on minimal path vectors in(2)can increase enormously with the size of the system.In order to solve this problem,two ordering heuristics are proposed to reduce the size of MMDD.By ordering the components of minimal path vectors before feed them into MMDD generation,we are able to improve the efficiency of modeling and calculation using MMDD method.Empirical experiments show that the proposed heuristics can reduce the size of MMDD,thus reduce the complexity of evaluating the mission reliability of swarm system ultimately.