Real-time yaw rate prediction based on a non-linear model and feedback compensation for vehicle dynamics control

Abstract

The yaw rate is a key state for vehicle dynamics stability control. However, the time delay from the steering input to the yaw rate measured by the yaw rate sensor changes with the frequency of the steering input and can also be affected by the road friction and the vehicle speed. In order to reduce the time delay effect on the vehicle dynamics stability controller, a new comprehensive yaw rate prediction method is proposed. A seven-degree-of-freedom non-linear vehicle model is adopted as the prediction model. A quadratic polynomial extrapolation method is used to compute the future steering-wheel input in the prediction horizon. A model-based yaw rate prediction algorithm is combined with a feedback compensator to guarantee the robustness of the algorithm. Meanwhile, a linear extrapolation method is utilized to minimize the defects of model-based prediction since the simplified vehicle model cannot fully model the hysteresis or other non-linearities of the vehicle systems. Experimental tests on the vehicle show that the combined algorithm is able to predict the yaw rate about 100 ms earlier than the sensor measurement is. Simulations show that the vehicle stability control strategy based on the proposed comprehensive prediction method has a better performance than the traditional control strategy does.