UAV Path Planning Based on Enhanced PSO-GA

Path planning for unmanned aerial vehicles (UAV) is a key technology for UAV intelligent system in the aspect of model construction. In order to improve the rapidity and optimality of UAV path planning, we propose a hybrid approach for UAV path planning in 2D environment. First, an enhanced particle swarm optimization algorithm (EPSO) combine with genetic algorithm (GA) which named as EPSO-GA is utilized to obtain the initial paths of UAV. In EPSO-GA, a hybrid initialization of Q-learning and random initial solutions is adopted to find the better initial paths for the UAV, which improves the quality of initial paths and accelerates the convergence of the EPSO-GA. The acceleration coefficients of EPSO-GA are designed as adaptive ones by the fitness value to make full use of all particles and strengthen the global search ability of the algorithm. Finally, the effectiveness of the proposed algorithm is proved by the experiments of UAV path planning.