A problem reduction based memetic algorithm for the vehicle routing problem with discrete split deliveries and pickups

Abstract

Vehicle routing problem with discrete split deliveries and pickups demands (VRPDSPDP), which considers simultaneously split deliveries, pickups demands and discrete demands, has recently received increasing attention in the academic community due to their potential real-world applications in logistic operations and supply chain. In this paper, to solve efficiently this computationally challenging problem, we proposed a problem reduction based memetic algorithm (PRMA for short). The proposed PRMA algorithm consists mainly of a problem reduction method aiming to reduce the size of problem, a crossover operator to generate an offspring solution from two parent solutions selected randomly from the population, a split method to convert a sequence of pairs of demands to several routes, a local search method to improve locally the quality of solutions, and a population updating strategy. We conducted extensive computational experiments to assess the performance of algorithm based on 222 benchmark instances commonly used in the literature, and the computational results show that the proposed algorithm is very efficient and significantly outperforms the state-of-the-art algorithm in the literature. In particular, the proposed algorithm improves the best-known results for 139 out of 222 instances, while matching the best-known results for 77 instances.