Capacitated Vehicle Routing Problem—A New Clustering Approach Based on Hybridization of Adaptive Particle Swarm Optimization and Grey Wolf Optimization

Capacitated vehicle routing problem (CVRP) is a challenging combinatorial optimization research and has drawn considerable interests from many scientists. However, no standard method has been established yet to obtain optimal solutions for all standard problems. In this research, we propose a two-phase approach: An improved k-Means algorithm for the clustering phase and a hybrid meta-heuristic based on Adaptive Particle Swarm—PSO and Grey Wolf Optimization—GWO (APGWO) for the routing phase. Our approach gives results very close to the exact solutions and better than the original k-Means algorithm. And for the routing phase, our experimental results show highly competitive solutions compared with recent approaches using PSO and GWO on many of the benchmark datasets.