车辆侧倾状态估计的研究

In view of the problem that under complex driving conditions the rolling state of vehicle cannot be accurately measured and hence vehicle roll cannot be effectively controlled, vehicle coupling dynamics and fuzzy T-S theory are used to design a model-based algorithm for T-S state observer, and fulfill the real-time and effective observation of vehicle rolling state in complex conditions in this paper. Firstly, vehicle vertical and lateral kinematic models under the conditions of different road excitations and steering wheel angles are constructed. Then, with consideration of the strong nonlinear features of tire forces, a tire T-S model based on T-S theory is proposed and then verified by simulation with "magic formula". Next, based on tire T-S model and vehicle dynamics model, Bayesian theory is applied to the design of the algorithm for vehicle T-S observer. Finally, CarSim software is used to conduct simulation verification on the real-time estimation of vehicle roll angle and roll rate by vehicle T-S observer under standard-A/C road excitations. The results show that T-S observer can effectively estimate the rolling state of vehicle with the maximum standard deviation less than 12%.