Risk perception and the warning strategy based on safety potential field theory.

Benefiting from the rapid development of communication and intelligent vehicle technology in recent years, most traffic information is capable of being collected, processed, and transmitted to each vehicle through a connected and automated vehicles (CAVs) system. To meet the higher requirements of driving safety in CAVs environment, it is necessary to develop more effective safety evaluation indicators that combine all the traffic information received by the vehicle. To this end, this study proposes a novel methodology for risk perception and warning strategy based on safety potential field model to minimize driving risk in the CAVs environment. A dynamic safety potential field model was constructed to describe the spatial distribution of driving risk encountered by vehicles. This safety potential field model can comprehensively consider the impact of various types of traffic information on driving risk. And then, a novel driving risk indicator, named potential field indicator (PFI), was established to evaluate the level of driving risk. Finally, an early warning strategy was proposed to prevent accidents, whose performance was evaluated by several simulations carried out through SUMO simulator. The comparison with some classic risk indicators indicate that our proposed PFI can more accurately reflect the actual driving risk faced by vehicles under different vehicle motion states and thus is more suitable for driving risk assessment in the CAVs environment. It is expected that the findings in this study could be valuable in improving the performance of strategic decision-making in driver assistance systems in the CAVs environment.