Research On Mission Planning Methods For Aerial Recovery Of Unmanned Aerial Vehicle Swarms

Relying on technological advances and developments,unmanned aerial vehicles(UAVs)are increasingly used to perform a variety of military and civilian missions.However,a single UAV will be difficult to meet future application requirements because of the increasing complexity of mission scenarios.Thus,large-scale clustering has become the main development trend of UAVs in recent years.To maximize the advantages of UAV swarms in terms of low cost,low battle damage,high mobility,and high flexibility,aerial recovery technology will be an integral part.In the future,the “launch-work-recovery-relaunch” mission execution mode achieved by the mother aircraft can greatly improve the efficiency ratio and sortie efficiency of UAV swarms and enable UAV swarms to have long-range,low-cost,multi-wavelength,hierarchical mission execution capabilities.Based on the new requirements for autonomous control capability,this paper focuses on the aerial recovery process of UAV swarms and conducts the research on recovery mission planning methods.Firstly,the basic concepts of aerial recovery mission planning,including mother aircrafts,UAVs and recovery equipment,are introduced and the operation flow of the “launch-work-recovery-relaunch” mission execution mode with aerial recovery process is given.Then,a generic recovery trajectory planning,recovery task allocation and recovery task scheduling model is established based on the simplified physical models of the mother aircrafts and UAVs.At the same time,a quantitative analysis of the model complexity is presented.In addition,the characteristics of the aerial recovery mission planning problem are studied and the basic components of an aerial recovery mission planning system are given.For the recovery trajectory planning problem with expected range constraint in three-dimensional environment,a recovery trajectory planning framework is established based on the decoupled simplification approach.In the two-dimensional plane,the homotopic description method and forward kinematic description met hod of UAV recovery trajectory are firstly proposed.Based on the two tr ajectory description method,the recovery trajectory with curvature constraint are established.Then,two recovery trajectory homotopy construction patterns are introduced and a homotopic trajectory planning approach is proposed to generate recovery trajectory with expected range.In the altitude direction,the recovery trajectory is smoothed by B-spine curve based on the end points constraint and performance limitation constraint of the UAV.In order to solve the aerial recovery task scheduling problem with a single mother aircraft,a trajectory evaluation method based on the Dubins theory is proposed to predict the recovery time window of UAVs,and then a genetic algorithm is used to optimize the optimal recovery sequence in the recovery task scheduling process.Then,the recovery time,recovery trajectory range and recovery location of each UAV are obtained by iteration to complete the recovery trajectory planning process.Meanwhile,a rescheduling strategy based on the genetic algorithm is proposed to deal with the dynamic events occurring in the recovery task scheduling process;a mother aircraft trajectory planning method based on the clustering algorithm,the greedy strategy and the Dubins theory is proposed to deal with recovery task scheduling problem without a priori mother aircraft trajectory.As a result,the adaptation of the recovery task scheduling method to complex scenarios is achieved.Considering the coupling mechanism between recovery task allocation,recovery task scheduling and recovery trajectory planning,a multi-stage aerial recovery mission planning framework is established in order to solve the aerial recovery mission problem with multiple mother aircrafts.Then,a hierarchical optimization algorithm and an improved particle swarm algorithm are proposed to realize the aerial recovery mission planning process.Finally,the effectiveness and feasibility of the two algorithms are verified through comparison simulation based on the multi-stage aerial recovery mission planning framework.