• 车辆工程 •

### 基于道路误差动力学模型智能车辆横向切换LMIs/H∞转向控制策略

1. 1.合肥工业大学，合肥， 230009
2.安徽理工大学，淮南，232001
• 出版日期:2015-08-25 发布日期:2015-08-25
• 基金资助:
国家自然科学基金资助项目(51075112，51405004)
National Natural Science Foundation of China（No. 51075112，51405004）

### Lateral Switched LMIs/H∞ Steering Control Strategy of Intelligent Vehicle Based on Path Error Dynamics Model

Wang Xuanyao1,2;Wang Qidong1,2

1. 1.Hefei University of Technology,Hefei,230009
2.Anhui University of Science and Technology,Huainan,Anhui,232001
• Online:2015-08-25 Published:2015-08-25
• Supported by:
National Natural Science Foundation of China（No. 51075112，51405004）

Abstract:

Based on the path error dynamics model,a lateral switched steering controller was designed for an uncertain linear model of the vehicle, which included switched PIDey control of the distance from the centerline of the lane to the center of mass of the vehicle, and switched PDeψ control of the direction errors from the lane to the vehicle. Setting the vehicle theoretic yawing angular velocity as the disturbance, the optimal solution of convex optimization problems of linear objective function, and the corresponding minimum disturbance rejection γ were acquired for the control strategy, which the state feedback γ-suboptimum H∞ norm and linear matrix inequalities(LMIs) were used. The road model of Alt3 from FHWA  was acquired from CarSim, and the reference velocity was obtained from the vehicle model D-Class/Sedan. In order to simplify the model, a sub-model with time-varying velocity was selected as the switching signal for the lateral switched steering controller. On the road with low adhesion coefficient, there  were both verified with 10 degrees of freedom vehicle dynamics model, which was  based on the path error dynamics model of the lateral control law  and have higher path tracking accuracy and stability than L2WVM(linear two-wheel vehicle model). With the changed structure parameters of vehicle model, it is verified that the  switched H∞ controller based on the path error dynamics model has better robustness than the L2WVM.