【摘要】 针对考虑电量限制的电动车辆在同时取送货场景下的路径规划问题，提出一种改进的文化基因算法.该算法采用基于极坐标的初始化方法生成初始解，引入一种最优分割方法优化解的质量，并设计多种特定的局部搜索算子提升搜索效率.同时，通过扰动机制避免陷入局部最优.将所提局部算子集成到其他算法中，在小、中、大规模数据集上进行对比实验，并结合Wilcoxon符号秩检验对不同算法进行性能分析.结果表明，该算法在各项评价指标上均优于对比算法，且提出的局部搜索算子加入后算法性能显著提升，验证了其在求解优质解方面的有效性.更多还原

【Abstract】 An improved memetic algorithm was proposed to address the electric vehicle routing problem with simultaneous pickup and delivery under battery capacity constraints.The algorithm employed a polar coordinate-based initialization method to generate initial solutions and introduced an optimal splitting method to enhance solution quality.Multiple specialized local search operators were designed to improve search efficiency,and a perturbation mechanism was incorporated to escape local optima.The proposed local search operators were integrated into other algorithms,and comparative experiments were conducted on small-,medium-,and largescale datasets.Performance evaluations were performed using the Wilcoxon signed-rank test.The results demonstrate that the proposed algorithm outperforms comparative algorithms across all evaluation metrics.Additionally,the performance of other algorithms significantly improves after incorporating the proposed local search operators,confirming their effectiveness in obtaining high-quality solutions.更多还原