A Stable Vehicle Platooning Scheme Based on Multi-Armed Bandits

Le résumé fourni par la source

Vehicle platooning technology enhances road utilization efficiency, reduces fuel consumption, and improves driving safety. However, conventional control approaches encounter significant stability challenges, including delays arising from centralized optimization based on global information exchange, rigid control policies that fail to account for vehicle heterogeneity, and the lack of effective incentive mechanisms, which collectively contribute to frequent protocol violations. To address these limitations, this paper proposes a stable platooning resource allocation framework that integrates multi-armed bandit (MAB) learning with incentive compatible mechanisms. A two-layer MAB architecture is developed to enable road segment selection and vehicle role assignment. The upper confidence bound (UCB) strategy is employed to balance exploration and exploitation, thereby improving decision robustness and system stability. Additionally, blockchain technology is leveraged to implement a transparent and tamper-resistant reward and penalty mechanism designed to discourage non-compliant behavior. Theoretical analysis shows that the proposed scheme enables stable platooning by aligning incentives through audits, controlling the budget for transfer settlements and using finite-horizon UCB learning behaviors. Extensive experiments illustrate that the proposal ensures equitable benefit distribution between leader and follower vehicles, exhibits strong resilience against malicious disruptions, and maintains acceptable computation overhead in large-scale scenarios.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.