基于系统辨识的车辆动力学建模方法

Vehicle dynamics modeling method based on system identification

针对车辆动力学理论建模过程中由于简化假设导致模型精度不高,进而影响车辆稳定性控制分析的问题,提出了一种基于系统辨识的车辆动力学建模方法,该方法综合了道路几何线形特征与实际系统中的未建模动态,在对理论模型分析的基础上,利用车载系统测量的输入输出数据以及最大信息准则确定候选模型结构,通过对理论模型的辨识修正,减小了建模误差。最后借助车载实验平台,针对车辆蛇形、转弯等不同工况进行了实验。结果表明,与理论建模方法相比,所建立的辨识模型更接近现实系统,模型的输出结果与实际测量值吻合较好,进而验证了该方法的有效性以及所建模型的适用性。

Aiming at the problem that present modeling methods based on vehicle dynamics theory cause low model accuracy due to simplified assumptions and influence the control analysis of vehicle stability,a vehicle dynamics modeling method based on system identification is proposed,which considers the geometrical characteristics of the road and the unmodeled dynamics in the actual system. Based on analyzing the established theoretical model,the structure of the candidate model is determined with the input and output data measured in the vehicle system and the Akaike＇s Information Criterion（ AIC）. The modeling error is reduced through the identification correction of the theoretical model. The slalom test widely used in the study of vehicle handling performance was conducted on the vehicle test platform for different operation conditions such as right turning,left U turning and etc. The results show that compared with the theoretical modeling method,the proposed identification model is closer to the real system; the outputs from the proposed identification model are consistent with the tested values,which verifies the effectiveness and usefulness of the proposed identification model.