With the popularity of electrification and driver assistance systems on vehicle dynamics and controls, the steering performance of the vehicle put forward higher requirements. Thus, the steer-by-wire technology is becoming particularly important. Through specific control algorithm, the steer-by-wire system electronic control unit can receive signals from other sensors on the vehicle, realize the personalized vehicle dynamics control on the basis of understanding the driver’s intention, and grasp the vehicle movement state. At the same time, to make these driver assistance systems better cooperate with human drivers, reduce system frequent false warning, full consideration of mutual adaptation for the systems and the driver’s characteristics is critical.
This paper focuses on the steering performance of steer-by-wire vehicle. Feature parameters are obtained from the virtual turning experiment designed on the driving simulator experimental platform. The identification model of driver steering behavior characteristics is established based on the experiment data with Back Propagation neural network as the aid of pattern recognition theory. The model is able to predict human driving behaviors and distinguish among different drivers, to classify the steering behavior. On this basis, according to different types of steering behavior, the variable steering angle ratio of steer-by-wire vehicle is designed to meet the individual needs of different driving habits. Finally, the hardware-in-the-loop simulation experiment validation shows that the proposed personalized variable steering angle ratio control strategy can meet the different steering preferences of the driver, and realize the angle ratio self-adaptive adjustment at high and low speed, improve the vehicle low-speed steering sensitivity, high-speed handling stability and active safety. This control strategy effectively enhances the steering performance and overall dynamics characteristics of steer-by-wire vehicles, achieving the human driver and vehicle coordination control.