Vehicle Crash Mitigation Strategy in Unavoidable Collision Scenarios: Focusing on Motion Planning by Considering a Generalized Crash Severity Model

Abstract

Motion planning in dynamic environment is crucial to the automated driving safety. In extremely emergency scenarios with unavoidable collisions, especially those with complex impact patterns, the potential crash risk should be well considered in motion planning. This paper proposes a motion planning algorithm for unavoidable collisions, which directly embeds a generalized crash severity index model to vehicle-to-vehicle collisions of multiple impact patterns. Firstly, the clothoid curve is used to sample the vehicle trajectory before collision, and a two-degree-of-freedom model is adopted to predict the vehicle poses corresponding to each sample path. Then, the crash severity index model is to estimate the potential crash severity of all sample paths. To improve the inferring time efficiency, a neural network is constructed and deployed to approximate the nonlinear severity model. Finally, the crash-severity-optimal trajectory is tracked through model predictive control method. Results show that by combining the braking and steering interventions for better crash severity reduction, the proposed strategy can achieve better mitigation effects than commonly-used collision-avoidance strategies. The deployment of real car experiment and sensitivity analysis demonstrate that the planning algorithm can guarantee real-time and reliably safe performances.