Study on hybrid control based steering system for intelligent vehicle in high-speed condition

Abstract

Through the analysis of the main types of accidents and three basic driving behavior in high-speed condition, it was concluded that steering system has both discrete state and continuous state. And then a multi-mode-switching control strategy was proposed, which used time to line crossing (TLC) and safe distance as switching rules. The damping, lane-keeping and lane-changing controller were designed respectively and were implemented on steer-by-wire system. In lane-keeping model, a new method is introduced to control the first derivative and the second derivative of the lateral offset of the preview point, which can not only prevent the excessive overshoot, but also reduce the time for the vehicle to return to the central line of the lane. The lane-changing model used the model predictive control (MPC) with multiple constraints to improve the stability of vehicle lane changing process. Based on the stability theory of hybrid system, a simple but effective stability supervisor was designed to prevent vehicle from the loss of control. Finally, the proposed control method is implemented on CarSim/Simulink and hardware-in-loop test platform, and the results proved the effectiveness of this strategy, and the stability of the switching process.