基于改进金豺优化算法的机器人路径规划

Robot Path Planning Based on Improved Golden Jackal Optimization Algorithm

针对传统金豺优化算法GJO求解移动机器人路径规划RPP问题时存在寻优能力差,易陷入局部最优的缺点,本文提出一种改进的金豺优化算法IGJO。IGJO算法引入反向学习机制构建初始金豺种群,以提升初始解的质量;使用改进非线性能量逃逸因子,避免迭代后期过早收敛。同时,算法融合了个体记忆功能的精英反向学习策略搜索当前种群优秀解的反向空间,以增强算法的勘探能力。最后,IGJO对比5种优化算法的实验数据,实验结果表明,IGJO算法均优于其他对比算法。

Aiming at the shortcomings of the traditional golden Jackal optimization algorithm GJO, such as poor optimization ability and easy to fall into local optimum when solving the RPP problem of mobile robot path planning, this paper proposes an improved golden Jackal optimization algorithm IGJO. IGJO algorithm introduced the opposition-based learning mechanism to construct the initial golden jackal population to improve the quality of the initial solution. The improved nonlinear energy escape factor was used to avoid premature convergence in the later iteration. At the same time, the elite opposition-based learning strategy integrated with individual memory function was used to search the reverse space of excellent solutions in the current population to enhance the exploration ability of the algorithm. Finally, IGJO compares the experimental data of five optimization algorithms. The experimental results show that IGJO algorithm is superior to other comparison algorithms.