The capacitated vehicle routing problem (CVRP) is a well-known optimization issue in transportation logistics. As a typical representative of swarm intelligence algorithm, antcolonyoptimization (ACO) has shown encou...
详细信息
The capacitated vehicle routing problem (CVRP) is a well-known optimization issue in transportation logistics. As a typical representative of swarm intelligence algorithm, antcolonyoptimization (ACO) has shown encouraging outcomes in CVRP. In contrast, ACO has limitations such as undesirable solutions and susceptibility to getting stuck in local optima. To address these challenges, a multi-strategy adaptive antcolonyoptimization with the k-meansclusteringalgorithm (kMACO) is proposed for solving CVRP in this study. In the initial stage of kMACO, k-meansclusteringalgorithm is introduced to enhance the quality of the initial solution. Simultaneously, a path-saving factor is added to the state transition rules to improve the success rate of planning. Moreover, the algorithm's global search capability is further enhanced by dynamically adjusting the pheromone volatilization coefficient. Then, a problem-specific crossover operator and three-stage local operators are designed to strike a balance between the global optimization and local search of kMACO. Finally, to confirm the effectiveness of kMACO, simulation experiments are conducted on three types of datasets. Compared with ACO and six other intelligent algorithms, the kMACO achieves the best-known solution in 17, 12, and 10 instances in benchmark sets A, B, and P, respectively.
暂无评论