【摘要】 标准蝙蝠算法在自动引导车(Automated Guided Vehicle, AGV)路径规划时寻优精度低、初期收敛速度慢且易陷入局部最优解，为此，本文提出一种改进的蝙蝠算法。首先，引入自适应动态惯性权重平衡算法的全局搜索与局部搜索；然后，利用差分进化算法的变异机制的交叉、选择操作择优保留后代增强算法迭代后期逃离局部最优区域的能力并优化求解精度；最后，通过分阶段搜索策略进行全维和单维搜索最优个体，提高算法的收敛效率。使用标准测试函数对改进蝙蝠算法进行收敛曲线和运行时间测试，并在相同环境下与标准蝙蝠算法进行仿真实施对比。实验结果表明，改进后的算法比标准蝙蝠算法收敛速度提升了30%,路径长度缩短16.8%且提高了寻优精度，具备良好的实用价值与应用潜力。更多还原

【Abstract】 The standard Bat Algorithm shows some limitations in Automated Guided Vehicle(AGV) path planning, include low optimization accuracy, slow initial convergence speed, and a tendency to fall into local optima. An improved bat algorithm is therefore proposed. First, an adaptive dynamic inertia weight is introduced. This balances the global search and local search capabilities of the algorithm. Second, the mutation mechanism of the differential evolution algorithm is utilized. Its crossover and selection operations help preserve superior offspring. This enhances the ability to escape local optima during later iterations and improves solution accuracy. Finally, a phased search strategy is adopted. It conducts both full-dimensional and single-dimensional searches for the optimal individual, which increases the convergence efficiency of the algorithm. The improved bat algorithm is tested using standard test functions. Convergence curves and running time are analyzed. Simulations are also performed under identical conditions for comparison with the standard bat algorithm. Experimental results verify that the improved algorithm achieves approximately a 30% faster convergence speed and a 16.8% shorter path length. Furthermore, it has improved the optimization accuracy. The improved algorithm shows good practical value and application potential.更多还原