TLHSA and SACA: two heuristic algorithms for two variant VRP models

Abstract

Vehicle routing problem (VRP) is a classical combinatorial optimization problem. Under this problem, we focus on two variant models which better capture the real-world scenes: multiple-orders pickup and delivery problem with time-bound window (MOPDPTW) and dynamic vehicle routing problem with time window (DVRPTW). To tackle MOPDPTW, a two-layers heuristic search algorithm is proposed. The inner layer of proposed algorithm searches possible solutions in global and sends them to the outer layer to find local optimal solution. In order to solve DVRPTW, a general dynamical algorithm framework is designed to tackle the dynamic nature of the problem. Then based on ant colony algorithm, we propose several effective strategies called pheromone preserving mechanism, pheromone updating based on important solution components and parameter self-adaptive adjustment, aiming to improve the solution construction process by ants. We validate our two algorithms on different standard benchmarks and the results indicate that our proposed algorithms are competitive and effective compared with the state-of-the-art approaches.