Research And Simulation On UAV Path Planning

Unmanned aerial vehicles(UAVs)are widely utilized across a diverse range of industries owing to their characteristics such as low cost,compact size,high agility,and ease of operation.However,the efficacy of UAV task completion primarily depends on trajectory planning,which must ensure efficient task execution and avoidance of potential threat areas.To achieve optimal UAV trajectory planning outcomes,this study mathematically models two-dimensional and three-dimensional UAV task environments using the grid method.Furthermore,this study proposes two novel algorithms,namely the jump-point search improved ant colony(JPIACO)algorithm and the multi-point sampling and back-end optimization strategy improved rapidly exploring random tree(MQB-RRT*)algorithm.The main research content of this paper is as follows:This paper proposes the JPIACO algorithm to address the issues of ant colony optimization(ACO)in searching for optimal two-dimensional UAV trajectories,including susceptibility to local optima and slow convergence.To improve convergence speed,the JPIACO algorithm transforms the initial path generated by the Jump Point Search(JPS)algorithm into the initial pheromone concentration of the ACO algorithm.Additionally,the algorithm introduces the turning cost factor into the heuristic function of the ACO algorithm to enhance the smoothness of the search path.Furthermore,the reward and punishment factors of the Wolf Pack Algorithm are introduced to adaptively adjust the iteration rule of the pheromone concentration at each iteration,increasing convergence speed and reducing the likelihood of the algorithm being trapped in local optima.Finally,through simulations,the superiority of the JPIACO algorithm over the improved Fireworks-ACO hybrid algorithm(APFWA-ACA)and the improved deadlock ant colony algorithm is verified.Furthermore,this paper proposes the MQB-RRT* algorithm based on threedimensional grid map to address the real-time deficiencies and low path quality of sampling-based algorithms in online three-dimensional trajectory planning for unmanned aerial vehicles.The algorithm first introduces a multi-point sampling strategy to improve the efficiency of initial path search by incorporating heuristic ideas into the sampling process of the Quick rapidly exploring random tree(Q-RRT*)algorithm,and then performs back-end optimization on each iteration path to enhance the search path accuracy.Finally,simulation results demonstrate that the MQB-RRT*algorithm can search for a suboptimal path from the starting point to the destination in a short time,and its generated path is faster and more accurate than the traditional QRRT* algorithm.This study presents new solutions for the global two-dimensional trajectory planning problem and the online three-dimensional trajectory planning problem of unmanned aerial vehicles by introducing the JPIACO algorithm and the MQB-RRT*algorithm,respectively.The proposed algorithms offer innovative strategies for UAV trajectory planning.