Multi-objective Cooperative Path Planning For Tracking Target By UAVs In Urban Environment

In recent years,unmanned aerial vehicle(UAV)has been widely used in military and civil fields for its advantages of low cost,high security and flexibility.Target tracking is one of the important applications of UAV.With the increasing attention paid to urban safety issues,in order to meet the needs of urban public security control and management,our country proposed the construction of Skynet project,and UAV could be an important supplement to maintain urban safety.Therefore,taking the urban environment as the background,this paper studies the path planning problem of UAVS tracking ground moving target.Considering the characteristics of tracking tasks,we use different modes to track the target in real time.Aiming at tracking the ground target for fast motion,UAVs carry out tracking task in persistent mode.To study the cooperative path planning of UAVs,we consider the occlusion of buildings to airborne sensor observation line of sight,UAVs control input cost,and sensor energy loss as performance indexes.According to the importance level requirement between performance indexes,different priorities are set.Combining with fuzzy optimization,priority constraints are planned by variable domain method.The distributed model predictive control method is used to deal with the maneuverability constraint,collision avoidance constraint between UAVs and collision constraint between UAV and buildings.Finally,the cooperative path planning model of UAV in persistent mode is constructed,and the cooperative path of UAVs is obtained by solving the optimization model online.Considering that in the actual tracking task,there are mostly unknown target,UAVs need to estimate and predict the target’s motion state.In the urban environment,the target is divided into two kinds of moving states,which are straight along the road and turning at the intersection.Considering the two different motion states,the motion state of target is quite different.The extended Kalman filter(EKF)and probability filter algorithm based on multi-model is proposed to estimate and predict the motion state of target.Combining the proposed filtering algorithm with the UAV cooperative path planning model for cooperative target,a cooperative UAV path planning model for non-cooperative target is constructed to complete the consistent tracking task.In order to track and protect ground target in low-speed motion,UAVs use standoff mode to track target.The cooperative path planning problem of UAV tracking cooperative target in urban environment is studied.Considering the maintenance of the distance between UAV and the target,the relative phase difference between UAVs and the control input cost as the performance index,the track optimization is carried out.According to the importance level requirement between performance indexes,priority relationships are set.Performance index functions are optimized by goal programming method and priority constraints is planned by the order of relaxation satisfaction method.In addition,the maneuverability constraint of UAV and collision avoidance constraint between UAVs are considered.Based on the distributed model predictive control method,a UAV cooperative path planning model in standoff mode is constructed by combining the UAV dynamic model,performance index functions and constraints.The optimization model is solved online to complete the cooperative path planning of UAVs in standoff mode.