Design of the Distribution Network for a “Collect-on- Delivery” Company in a Metropolitan Context using Simulated Annealing with Path Relinking

Parcel delivery service by collect-on-delivery (COD, a.k.a. cash-on-delivery) companies is a new distribution mode in China. The design of the distribution network of a COD company in urban areas is of great importance and a relatively new research direction. In this paper, we aim to model this distribution network as a two-echelon location routing problem (2E-LRP) and propose a novel efficient optimization algorithm by combining simulated annealing (SA) and path relinking (PR) to solve this NP combinatorial optimization problem. The experimental results are reported for 84 test problems from Nguyen’s, Prodhon’s, and Sterle’s benchmark. They indicate that the proposed SA-PR algorithm is competitive with other existing state-of-art algorithms for 2E-LRP from the perspectives of both quality of solution and computational time. Therefore, SA-PR is an effective metaheuristic to solve 2E-LRP problems, i.e., design distribution network of a COD company in a metropolitan context.