Abstract
Many of the algorithms for solving vehicle routing problems expose parameters that strongly influence the quality of obtained solutions and the performance of the algorithm.Finding good values for these parameters is a tedious task that requires experimentation and experience. Therefore, methods that automate the process of algorithm configuration have received growing attention. In this paper, we present a comprehensive study to critically evaluate and compare the capabilities and suitability of seven state-of-the-art methods in configuring vehicle routing metaheuristics. The configuration target is the solution quality of eight metaheuristics solving two vehicle routing problem variants. We show that the automatic algorithm configuration methods find good parameters for the vehicle route optimization metaheuristics and clearly improve the solutions obtained over default parameters. Our comparison shows that despite some observable differences in configured performance there is no single configuration method that always outperforms the others. However, largest gains in performance can be made by carefully selecting the right configurator. The findings of this paper may give insights on how to effectively choose and extend automatic parameter configuration methods and how to use them to improve vehicle routing solver performance.
Highlights
The vehicle routing problem (VRP) is a practical, relevant, and challenging problem that has been extensively studied by the artificial intelligence (AI) and operations research (OR) communities
We focus on two variants: the capacitated vehicle routing problem (CVRP) and the vehicle routing problem with stochastic demands (VRPSD)
We focus on seven state-of-the-art algorithm configuration methods: CMA-ES [21, 64], GGA [1], Iterated F-Race [3], ParamILS [30], REVAC [46], SMAC [27], and uniform random sampling (URS) [64]
Summary
The vehicle routing problem (VRP) is a practical, relevant, and challenging problem that has been extensively studied by the artificial intelligence (AI) and operations research (OR) communities. In the field of vehicle routing research, Pellegrini and Birattari [48] compared the performance of different metaheuristics with and without automatic algorithm configuration and concluded that, in every instance, the automatically configured version of the solution algorithm yielded better results than the corresponding non-configured one. Besides our preliminary work presented in [52], there is no comprehensive comparative study on automatic algorithm configuration of vehicle routing solvers. This study addresses this knowledge gap by investigating the performance of recent automatic configuration methods in the domain of routing algorithms. 3. How does the performance of configurators vary with different metaheuristics, vehicle routing variants, and problem instances?
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have