Mission-Oriented Reliability Modeling And Prediction For Unmanned Aerial Vehicle Swarm

Unmanned combat has become the trend of future warfare,and as an important part of unmanned combat,UAV swarm has attracted the attention of military powers in recent years.With the increasing scale of UAV swamr,the reliability related to UAV swarm emerge.Unlike manned platforms,the reliability of UAV swarm shows the characteristics of changing from "platform-centered" to "mission-centered".Therefore,in view of the characteristics of UAV swarm operations,it is necessary to integrate the reliability modeling tools and evaluation methods closely with mission conditions in order to meet the requirements of cluster reliability prediction and provide theoretical support for battlefield decision-making.In the existing research,voting model is the basic idea of UAV swarm modeling,but this method can not reflect the actual technical constraints and mission requirements.For example,because of the communication requirement of UAV,the distance between each UAV can not be too long,because of the requirements of reconnaissance tasks,UAV swarm needs to ensure that the target area can not appear reconnaissance blind spots.For this reason,this paper takes the engineering requirement of mission reliability modeling and analysis of UAV swarm as the starting point,and based on the phased mission system and consecutive voting model,aiming at the problem of mission reliability modeling and evaluation of UAV swarm,studies the comprehensive solution method of phased mission system and consecutive voting model.The mission reliability modeling and fast prediction of UAV swarm are realized.The main contents of this paper are as follows:1)A method is proposed to analyze the reliability modeling and prediction for unmanned aerial vehicle swarm based on the actual situation.The existing theory constructs the UAV swarm as a voting model.However,using voting model to model UAV swarm under specific mission conditions can not reflect the real situation.Considering the engineering requirement and mission requirement of UAV swarm,the mission reliability of UAV swarm is modeled and predicted based on consecutive voting model,which makes the mission reliability more close to the real situation.2)A comprehensive analysis method of one-dimension consecutive voting model and phased mission system is proposedThe one-dimensional consecutive voting model and phased mission system are deeply studied.According to the event decomposition method of one-dimensional consecutive voting model and the combination of the two models,a recursive formula is proposed to solve the problem.The correctness of the formula is proved by comparing with the simulation analysis of enumeration method,and the calculation speed is greatly improved.On this basis,according to the characteristics of the phased mission system,an optimization formula is proposed to further optimize the computational efficiency of the model.3)A comprehensive analysis method of two-dimensional consecutive voting model and phased mission system is proposed.On the basis of one-dimensional consecutive voting model,the problem of combining two-dimensional consecutive voting model with phased mission system is studied.Based on the idea of Y-M algorithm,a recursive formula is proposed to solve this problem,and the mathematical analytic method of this kind of problem is obtained.The mission reliability modeling and evaluation of UAV swarm is carried out based on the proposed model.The correctness of the formula is proved by simulation analysis and comparison with enumeration method.4)A method is proposed to improved the two-dimensional consecutive voting model for UAV swarm modeling,and put forward the optimization model for the actual situation.Through the study of two-dimensional consecutive voting model and the improvement of function in the formula,the UAV swarm modeling based on twodimensional consecutive voting model is more close to the actual situation,and the obstacle problem,formation problem and multi-type UAV cooperative combat problem in the course of executing mission are solved.Based on the importance theory,two planning methods are proposed to improve the reliability of UAV swarm,which improves the adaptability of the model and realizes the reliability modeling and fast prediction of UAV swarm mission.