AUTHOR=Li Huilan , Wan Yun , Shang Junning , Xu Xiangyang 

TITLE=Risk-aware local traffic safety evaluation for autonomous new energy vehicles based on virtual force modeling

JOURNAL=Frontiers in Mechanical Engineering

VOLUME=Volume 11 - 2025

YEAR=2025

URL=https://www.frontiersin.org/journals/mechanical-engineering/articles/10.3389/fmech.2025.1658915

DOI=10.3389/fmech.2025.1658915

ISSN=2297-3079

ABSTRACT=IntroductionEnsuring safe driving and effective risk control is critical for the dynamic operation of autonomous new energy vehicles (NEVs), particularly under complex traffic conditions. A key challenge lies in unifying risk assessment across diverse driving scenarios, which hinders reliable and adaptive risk management in autonomous systems.MethodsTo address this challenge, a local traffic risk evaluation framework tailored for NEVs is proposed, grounded in the principle of least action. The method constructs a virtual force system centered on the autonomous NEV, integrating three components: (1) virtual risk force to capture vehicle-to-vehicle interaction risks; (2) virtual driving force reflecting the vehicle’s motion intention; and (3) virtual regulatory force to enforce traffic rule compliance. By modeling the action of this force system, a novel metric for local traffic safety is formulated, enabling real-time risk assessment and informing control strategies.ResultsThe proposed method was validated through simulations across typical hazardous scenarios, including rear-end collisions, emergency deceleration, lane changes, and intersection conflicts. Simulations demonstrated that the framework enables timely risk perception and adaptive control behaviors (e.g., braking, evasive lane changes), which substantially improved the driving safety of NEVs.DiscussionThis work provides a unified and computationally efficient tool for enhancing risk-aware decision-making and control in autonomous NEVs. By addressing the challenge of scenario-unified risk assessment, it contributes to the safer deployment of autonomous NEVs and their more intelligent integration into traffic systems.