Experimental evaluation of collision avoidance system using state dependent Riccati equation technique

In emergency cases a vehicle can arrive in a situation where it has to avoid colliding with one or more obstacles. At the same time the vehicle needs to stay within road boundaries, satisfy acceleration and jerk limits, fulfil stability requirements and respect that the vehicle response is very much depending on the vehicle speed and tire-road friction. The real time solution that covers the full scope of such a multi-variable problem is difficult. Currently various approaches that address only a limited set, which usually relax the problem, have been proposed until now. In this study, the development and experimental evaluation of a Collision Avoidance System is presented which is based on the State Dependent Ricatti Equation (SDRE) control technique. The proposed controller is based on a vehicle model with nonlinear tire characteristics which can account for speed and road friction influences. An extended linearization scheme of the system's state space equations on the basis of the combined slip Pacejka tire model is used. On-line control regulation generates the steering angle at each time increment by solving efficiently an Algebraic Ricatti Equation. The proposed method is implemented and tested on a TOYOTA Prius test vehicle with an active steering system. The results show the effectiveness of the proposed controller in smooth and safe collision avoidance, keeping within the boundaries of stability that are dictated by the road friction conditions.