最小二乘支持向量机在汽车动态系统辨识中的应用

Application of LS-SVMs in the Automobile Dynamical System Identification

汽车转向时动态系统参考模型对于汽车稳定性的控制有重要影响.基于最小二乘支持向量机算法,应用网络搜索和交叉验证的方法选择支持向量机参数,并将其应用于汽车转向时的非线性动态系统辨识,取得了良好的辨识效果,建立的参考模型能够较充分地描述汽车动力学行为.

The dynamical desired model has an important impact on the vehicle stability control during cornering. A nonlinear modeling method based on Least Squares Support Vector Machines (LS-SVMs) was proposed. LS-SVMs hyperparameters are optimized with grid-search and cross-validation methods. LS-SVMs is applied to modeling the desired vehicle dynamical system. The research result shows that it has good performance on the identification of nonlinear dynamical system. The desired model can fully represent the automobile dynamics.