Abstract: An adaptive chaotic particle swarm optimization algorithm（SACPSO） is proposed for three-dimensional space path planning. Firstly, the three-dimensional space environment modeling is carried out, and considers the three evaluation functions of path length, obstacle risk degree and path smoothness to formulate the fitness function. Then a new adaptive update strategy is proposed for the three control parameters in the algorithm, so as to dynamically adjust the global exploration and local exploitation capabilities of the algorithm. Finally, when the population falls into the local extremum, the proposed adaptive logistic chaotic map is used to optimize the global optimal particle and guide the population to jump out of the local extremum point. Comparing the algorithm with other improved particle swarm optimization algorithms, the results show that the algorithm uses fewer iterations when converging to the global optimal solution, and the quality of the generated path is higher, which effectively improves the computational efficiency and reliability of particle swarm optimization used in path planning problem in three-dimensional space.

Key words: path planning, three-dimensional space, particle swarm, self-adaptive, Logistic chaotic map

摘要： 提出一种自适应混沌粒子群优化算法（SACPSO）用于三维空间路径规划。首先进行三维空间环境建模，并考虑使用路径长度、障碍物危险程度和路径平滑度三个评价函数来制定适应度函数；然后对算法中的三个控制参数提出了一种新的自适应更新策略，以此来动态调整算法的全局探索和局部开发能力；最后当种群陷入局部极值时，利用提出的自适应Logistic混沌映射对全局最优粒子进行混沌优化，引导种群跳出局部极值点。将该算法与其他改进的粒子群算法比较，结果表明，该算法在收敛到全局最优解时所用迭代次数更少，生成路径质量更高，有效地提高了粒子群算法应用于三维空间路径规划时的计算效率和可靠性。

关键词: 路径规划, 三维空间, 粒子群, 自适应, Logistic混沌映射