Dynamic clustering for multi-depot capacitated vehicle routing with time windows: A CW heuristic approach

Abstract

Logistics and transport industries face challenges in efficiently planning routes to minimize costs. The Vehicle Routing Problem (VRP) aims to optimize vehicle routes for customer deliveries. This study explores a heuristic approach for solving the Multi-Depot Capacitated Vehicle Routing with Time Windows (MD-CVRPTW) problem, employing the Clarke-Wright (CW) Heuristic algorithm. A two-step model is presented, consisting of clustering followed by cluster optimization using the Clarke-Wright (CW) algorithm. The objective is to identify the optimal route from the warehouse, minimizing total transportation costs while adhering to capacity and time constraints and fulfilling customer location requirements. The optimization process is conducted on a dataset comprising 30 nodes, with the results demonstrating the algorithm's capability to generate satisfactory solutions.