The parallelization of a two-phase distributed hybrid ruin-and-recreate genetic algorithm for solving multi-objective vehicle routing problem with time windows

Solving multi-objective vehicle routing problem with time windows (MOVRPTW) involves satisfying two or more objectives. However, not many algorithms can achieve a broader set of Pareto optimal front and consistent solutions that have the least difference in magnitude. Therefore, it is rewarding to develop a coveted algorithm that solves these shortcomings. In this paper, we propose the parallelization of a two-phase distributed hybrid ruin-and-recreate genetic algorithm (HRRGA). The algorithms in HRRGA run in either the HRRGA and the hybrid ruin-and-recreate (HRR) phase or the HRR phase. In HRRGA phase, different HRR strategies are used with the hybrid genetic algorithm (HGA) in the near-optimal solution computation. However, in the HRR phase, only the HRR strategies are used. The algorithms in HRRGA are executed in parallel and harness its power of exploitation and exploration. These strategy combinations improve the diversity of the Pareto optimal front while delivering the generated solutions with the least difference in magnitude. Our experiment with Solomon's benchmark set shows that HRRGA has superior results compared to the recently published hybrid algorithm, has diverse set of Pareto optimal fronts, and achieves several novel Pareto optimal fronts compared to the best-known solutions.