Improved Particle Swarm Path Planning Algorithm Based on UAV

This study addresses the challenges encountered by traditional multi-agent pathfinding methods in the applications of fixed-wing UAVs, including inadequate adaptation to dynamic constraints, lack of energy optimization, low exploration efficiency, and inefficient data utilization. To tackle these issues, we propose an improved independent Q-learning algorithm that incorporates the dynamics of fixed-wing UAVs. Initially, we expand the traditional multi-agent pathfinding action space and incorporate constraints to develop a model that is compatible with the motion characteristics of fixed-wing UAVs. Next, we develop a multi-dimensional reward function system based on independent Q-learning. Additionally, we introduce a heuristic policy that utilizes directional weights, effectively balancing exploration efficiency and algorithm convergence speed. Finally, we implement a shared experience replay buffer and a reward reconstruction mechanism to enable multi-agent experience sharing and collaborative optimization. Experimental results indicate that our proposed algorithm surpasses the A* algorithm in total path energy consumption. Furthermore, the heuristic policy demonstrates a twofold improvement in convergence speed and the success rate of path planning compared to the $ \varepsilon $-greedy policy. The shared experience replay buffer also shows enhanced performance in convergence speed.

With the rapid development of Unmanned Aerial Vehicle (UAV) technology, the path planning problem in UAV cooperative operations has become increasingly prominent. This paper studies the path planning problem of UAV swarms, aiming to propose an effective method to improve the efficiency and safety of task execution. The Marine Predator Algorithm (MPA) is a popular optimization algorithm known for its high performance. However, it sometimes encounters issues such as low efficiency in spiral flight and slow predator speed, leading to slow convergence and susceptibility to local optima. To address these limitations, this paper proposes the Multi-strategy Enhanced Marine Predator Algorithm (MEMPA). The four enhancement strategies are: random spiral flight strategy, crisscross strategy, centroid boundary control strategy, and improved update rules for eddy formation and the FADs’ effect. These strategies collectively enhance the algorithm’s global search ability and its capacity to escape local optima. Compared with FDB-PPSO, AGWO, HPHHO, LSHADE, LSHADE-cnEpSin, MadDE, NMPA, QQLMPA, and the original MPA across four different dimensions of the 30 CEC2017 test functions, the Wilcoxon rank sum test results show that MEMPA outperforms the original MPA by 21, 19, 22, and 25 on the four dimensions, respectively. According to the results of the Friedman mean rank test, the top three algorithms are LSHADE, MPA, and MEMPA, with scores of 2.91, 2.77, and 2.14, respectively. Additionally, the proposed MEMPA was used to solve the UAV swarm path planning problem. The results show that the solutions obtained by MEMPA are 10%, 10%, and 13% better than those obtained by MPA, NMPA, and QQLMPA, respectively. It is superior to the selected comparison algorithms in terms of solution effectiveness and stability, demonstrating its potential in practical applications.

This paper addresses the issues of target unreachability and local optima in traditional artificial potential field (APF) methods for UAV swarm path planning by proposing an improved collaborative obstacle avoidance algorithm. By introducing a virtual target position function to reconstruct the repulsive field model, the repulsive force exponentially decays as the UAV approaches the target, effectively resolving the problem where excessive obstacle repulsion prevents UAVs from reaching the goal. Additionally, we design a dynamic virtual target point generation mechanism based on mechanical state detection to automatically create temporary target points when UAVs are trapped in local optima, thereby breaking force equilibrium. For multi-UAV collaboration, intra-formation UAVs are treated as dynamic obstacles, and a 3D repulsive field model is established to avoid local optima in planar scenarios. Combined with a leader–follower control strategy, a hybrid potential field position controller is designed to enable rapid formation reconfiguration post-obstacle avoidance. Simulation results demonstrate that the proposed improved APF method ensures safe obstacle avoidance and formation maintenance for UAV swarms in complex environments, significantly enhancing path planning reliability and effectiveness.

This paper presents an algorithm for trajectory planing using potential fields and swarms with vortex behavior. This algorithm is called Vortex Swarm Path Planning Algorithm (VSPP). There are several methods for trajectory planning in robotics, the most popular is the one based on artificial potential fields. However this method has the disadvantage of local minima which can make the robot fail when reaching the destination point. The proposed algorithm here presented employs a model of active Brownian particles with vortex behavior that allows the trajectory planning evading local minima.