多策略优化麻雀搜索算法及其路径规划的应用

Multi-strategy optimized sparrow search algorithm and its application in path planning

针对传统的麻雀搜索算法存在初始解质量较差、收敛效率低和求解精度不高等缺点,提出了一种基于多策略优化的麻雀搜索算法(LAFSSA)。首先,利用Logistic混沌映射初始化麻雀种群,丰富其多样性;其次,利用自适应惯性因子改进麻雀种群的生产者的位置更新方式,使得算法在迭代初期专注于全局空间的搜索,在迭代后期专注于局部区域的挖掘;最后,融合柯西变异算法进一步增强最差麻雀个体与最优麻雀个体的交流,加快算法的收敛速度,使其在迭代后期能够跳出局部最优。将LAFSSA应用于移动机器人的路径规划,仿真结果表明,与传统的麻雀搜索算法相比,LAFSSA不仅能够减少路径寻优时间,缩短移动机器人的路径长度,还能减少路径转折次数,提升路径的平滑性。

In view of the shortcomings of traditional sparrow search algorithm(SSA),such as poor initial solution quality,low convergence efficiency and low solution accuracy,a multi-strategy optimizd SSA is proposed,which is called LAFSSA.Firstly,the sparrow population is initialized by Logistic chaotic mapping to enrich its diversity.Secondly,the adaptive inertia factor is utilized to improve the location update method of producers of the sparrow population,so that the algorithm focuses on global space search at the beginning of iteration and on local area mining at the end of iteration.Finally,the Cauchy mutation algorithm is integrated to further enhance the communication between the worst sparrow individual and the best one,which accelerates the convergence speed of the algorithm and enables it to get rid of the local optimum in later iteration.The LAFSSA is employed to the path planning of the mobile robot,and experiment results show that,compared with the SSA,the proposed LAFSSA can not only reduce the path optimization time and shorten the path length of the mobile robot,but also decrease the number of path turns and improve the smoothness of the planned path.