Adaptive large neighborhood search for 2-echelon vehicle routing problem in last-mile delivery

This research studies the Multi-depot Two-Echelon Vehicle Routing Problem while considering route optimization from multi depots to certain intermediary facilities, i.e. satellites, to the end client (MD-TEVRP). The MD-TEVRP involves the delivery of goods from numerous depots to satellites and from satellites to ultimate consumers. The MD-TEVRP includes satellite time window limitations for satellites on the two-echelon network and tries to discover the most cost-effective paths for the two-echelon network. In this paper, a mixed integer programming model for the MD-TEVRP is developed and small-scale examples were solved by CPLEX solver. In large-scale problem, Adaptive Large Neighborhood Search (ALNS) heuristic is adapted to find solutions