A scalable learning approach for the capacitated vehicle routing problem
Year: 2024
Type: article
Abstract: Designing efficient heuristics for the different variants of vehicle routing problems and customising the heuristics to various input distributions is a time-consuming and expensive task. In recent years, end-to-end machine learning techniques have been developed as they are easy to modify for differentproblem variants, thereby saving on the design time to develop new efficient heuristics. However, existing learning techniques struggle to provide highquality solutions on problem instances with hundreds to thousands of customers in a reasonable time. Furthermore, many of the end-to-end heuristicsalso do not guarantee that solutions obey fleet-size constraints. We propose a heuristic for solving large capacitated vehicle routing problem (CVRP) thatcarefully integrates a machine learning heuristic with Integer Linear Programming techniques. To address the issue of poor solutions generated byend-to-end machine learning approaches on larger instances, we dynamically partition the CVRP problem instance into smaller sub-problems and applya reinforcement learning heuristic on the smaller sub-problems. This allows the reinforcement learning heuristic to always operate on smaller problemssimilar in size to those for which it was trained. The reinforcement learning heuristic generates many solutions for each sub-problem which are then combined using a set partitioning approach based on a ILP formulation. The set partitioning ILP also guarantees that solutions obey fleet-size constraints.We evaluate the performance of our heuristic on a difficult set of benchmark instances with hundreds to thousands of nodes, achieving small gaps(less than 3% on average) with respect to best known solutions, significantly improving upon the solution quality of the existing learning heuristics.
Source: Computers & Operations Research
Institution: University College Dublin
Language: English
Subfield: Industrial and Manufacturing Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: hybrid
Estimated APC paid: $3,470
Funder: Science Foundation Ireland
OpenAlex ID: W4401433491