Research On The Algorithm Of Task Allocation And Path Planning For Multi-UAV

With the development of information technology and UAV(unmanned aerial vehicle)confrontation technology,as a new type of military weapon with high cost performance,UAV has a broader application prospect in modern war.Consequently,in order to meet the needs of actual battlefield environment and tasks diversity,while improving the completion rate and overall efficiency of tasks,multi-UAV task planning algorithms emerged.According to in-depth research and analysis,the current task planning algorithm cannot effectively meet the needs of task dynamics,multi-UAV coordination,and real-time planning,and in order to give full play to the independent decision-making and cooperative combat capability of multi-UAV,in this thesis,the main research contents of multi-UAV task allocation and path planning algorithms are as follows:(1)Overview of multi-UAV task planning algorithms.Based on the study of task allocation and path planning algorithms proposed in recent years,the characteristics and application scenarios of the related algorithms are analyzed,and the key problems at the current stage are pointed out.(2)Design of initial allocation and reallocation of multi-UAV tasks.Based on the analysis of task requirements,an allocation scheme which combines initial allocation and reallocation based on the CMTAP(cooperative multiple task assignment problem)model is proposed in this thesis.The scheme can be divided into three stages:(1)According to the CMTAP model and comprehensive consideration of terrain,threats,UAV performance,and other constraints,at the same time,meeting the requirements of task coordination and other requirements,a scientific and standardized allocation model is established.(2)Based on the allocation model and centralized solution method,an improved discrete particle swarm optimization algorithm is proposed to make full use of the computational performance of the centralized computing mode,optimize the task allocation results and improve the overall performance of task execution.(3)In response to emergencies such as drone damage during the task execution process,an auction algorithm based on the improved contract network model was proposed with the optimized bidding agent selection mechanism and contract type to meet the requirements of task redistribution problem and real-time.The simulation results prove that the scheme has certain advantages in solving multi-UAV task allocation,and can deal with emergencies to ensure the task execution capability.(3)A particle swarm potential field hybrid algorithm of path planning.According to the research of multi-UAV cooperative path planning and the complex dynamic battle environment,a particle swarm potential field hybrid algorithm was proposed with improved update formula and adaptive capacity of inertial weights and learning factors,while combining potential field force inspired information to enhance the convergence performance of the algorithm and avoid falling into a local optimum.The simulation results show that the hybrid algorithm can meet the safety and real-time requirements of path planning in static and dynamic scenarios,meanwhile,the stability and superiority of the algorithm are verified by comparing with similar algorithms.(4)Multi-UAV 3D simulation platform.Based on the analysis of battlefield environment and the combat process of UAV,the multi-UAV 3D simulation platform is established using My Eclipse and JMonkey Engine with some functions such as parameter setting,task planning and results playback.Meanwhile,the platform provides some relevant algorithm interfaces for subsequent expansion,and supports 3D and multi-angle scene display to facilitate observation and analysis of task planning results.Finally,the feasibility and good interactive performance of the platform is verified.