Problem
Research questions and friction points this paper is trying to address.
Electric Autonomous Dial-a-Ride Problem
Uncertainty
Large Neighborhood Search
Adaptive Operator Selection
Chance-constrained
Adaptive Operator Selection in Bilevel Large Neighborhood Search for Electric Autonomous Dial-a-Ride Problem under Uncertainty
🤖 AI Summary
This study addresses the electric autonomous vehicle on-demand mobility problem under uncertainty, where the repair phase of existing Adaptive Large Neighborhood Search (ALNS) methods suffers from strong dependence on insertion operator selection and frequent time-window constraint violations. We propose a bilevel ALNS framework integrating large language model (LLM)-based state awareness with performance-adaptive mechanisms, modeling stochastic variables via chance-constrained programming across both deterministic and chance-constrained variants. By systematically comparing six operator selection strategies, we reveal that no single strategy is universally optimal and introduce a novel paradigm for dynamic operator selection based on instance characteristics. Experimental results demonstrate that this strategy achieves robust performance on small-scale instances while exhibiting significant advantages in large-scale, highly constrained scenarios, effectively enhancing solution robustness in complex environments.
📝 Abstract
The electric autonomous dial-a-ride problem (EADARP) extends the classical dial-a-ride problem by incorporating battery and charging constraints for electric vehicles. In practice, travel-time uncertainty can cause violations of time-window constraints. Large neighborhood search is effective for solving the EADARP, but its performance can depend on the choice of insertion operator during the repair phase. This paper investigates insertion-operator selection within a bilevel large neighborhood search framework for deterministic and chance-constrained variants of the EADARP. In the chance-constrained variant, arc travel times are modeled as independent normally distributed random variables, and upper time-window constraints are enforced probabilistically. We consider six selection methods, namely fixed greedy insertion, fixed regret-based insertion, random selection, a deterministic state-based rule, performance-adaptive ALNS selection, and LLM-based state-aware selection. Experimental results show comparable performance on smaller instances, while differences become more evident on larger and more constrained instances. There is no single strategy that performs best across all instances, and the relative performance of the LLM-based, rule-based, and ALNS strategies varies with the problem instance and experimental setting.
Innovation
Methods, ideas, or system contributions that make the work stand out.
Bilevel Large Neighborhood Search
Electric Autonomous Dial-a-Ride Problem
Chance-Constrained Optimization
Adaptive Operator Selection
LLM-based Selection
🔎 Similar Papers
No similar papers found.
