Instance-Aware Parameter Configuration in Bilevel Late Acceptance Hill Climbing for the Electric Capacitated Vehicle Routing Problem

2026-05-01

AI Summary

This study addresses the electric capacitated vehicle routing problem, where the performance of combinatorial optimization algorithms is highly sensitive to parameter settings. Due to structural, demand-related, and energy-consumption differences across problem instances, a single globally tuned parameter configuration often yields suboptimal results. To overcome this limitation, this work proposes an instance-aware parameter configuration mechanism integrated into a two-level late acceptance hill-climbing algorithm. The approach leverages offline tuning to generate instance-specific parameter labels, then combines feature extraction with regression modeling to predict near-optimal parameters for unseen instances prior to execution. Evaluated on the IEEE WCCI 2020 benchmark suite and extended test sets, the method achieves an average 0.28% reduction in objective values compared to global parameter tuning, translating into substantial cost savings—on the order of tens of millions—in large-scale transportation operations.

Abstract

Algorithm performance in combinatorial optimization is highly sensitive to parameter settings, while a single globally tuned configuration often fails to exploit the heterogeneity of instances. This limitation is particularly evident in the Electric Capacitated Vehicle Routing Problem, where instances differ in structure, demand patterns, and energy constraints. This paper investigates instance-aware parameter configuration for Bilevel Late Acceptance Hill Climbing, a state-of-the-art metaheuristic for the Electric Capacitated Vehicle Routing Problem. An offline tuning procedure is used to obtain instance-specific parameter labels, which are then mapped from instance features via a regression model to enable parameter prediction for unseen instances prior to execution. Experimental results on the IEEE WCCI 2020 benchmark and its extensions show that the proposed approach achieves an average objective value reduction of $0.28\%$ across eight held-out test instances relative to a globally tuned configuration. This corresponds to a significant cost reduction in multimillion-dollar transportation operations.

Problem

• Electric Capacitated Vehicle Routing Problem
• parameter configuration
• instance heterogeneity
• combinatorial optimization
• metaheuristic

Innovation

• instance-aware parameter configuration
• bilevel late acceptance hill climbing
• electric capacitated vehicle routing problem
• regression-based parameter prediction
• combinatorial optimization