Abstract: This paper is concerned with a novel Random Delayed Particle Swarm Optimization (RDPSO) algorithm which is proposed for the disadvantages of standard particle swarm optimization algorithm, such as fast convergence speed in the early stage, premature in the later period and local trapping phenomenon in the searching process. In this algorithm, the velocity update model switches from one mode to another according to the expectation of the random variable. Furthermore, in order to reduce the occurrence of local trapping phenomenon and expand the search space in the searching process, the random time-delays are introduced to the velocity updating equation. A simulation example is provided to verify that the integrated performance of the proposed algorithm is better than the other improved PSO algorithms. Finally, the RDPSO algorithm is applied to the UAV path planning in oilfield inspection. Experiments show that the RDPSO algorithm can simultaneously avoid the occurrence of local trapping phenomenon and ensure the convergence speed.

Key words: UAV, path planning, random delayed information, improved particle swarm optimization

摘要： 针对标准粒子群算法初期收敛速度快，后期容易陷入早熟，局部寻优等缺点，提出了一种新的随机时滞粒子群优化算法(Random Delayed Particle Swarm Optimization, RDPSO)。该算法中，速度更新模型将根据随机变量期望值从一种模式切换到另一种模式。此外，为了减少陷入局部最优现象的发生并扩大搜索空间，在速度更新方程中添加了随机时滞。仿真结果表明了所提出的RDPSO算法的综合性能优于其他改进的PSO算法。最后,将RDPSO算法应用于油田巡检无人机航迹规划问题，实验证明RDPSO算法有效地避免了局部最优现象的发生，同时保证了收敛速度。

关键词: 无人机, 航迹规划, 随机时滞, 改进粒子群算法