A Robust Guiding Torque Control Method for Automatic Steering Using LMI Algorithm

• Main Authors: Guangtong Bai, Chunjiang Bao, Jian Wu, Dexi Wu, Junda Zhang, Xiangkun He
• Format: Article
• Language: English
• Published: IEEE 2020-01-01
• Series: IEEE Access
• Subjects: Steering system; intelligent vehicle; LMI; human-machine co-driving
• Online Access: https://ieeexplore.ieee.org/document/8968403/

The existing path tracking methods usually neglect the effect of the drivers on the steering control. This paper proposes a robust steering control method of human-machine steering torque superposition based on linear matrix inequality (LMI) algorithm. First, the model for solving steering superposition torque introduces the steering system and steering resistance torque model in addition to the vehicle model, which increases the nonlinearity and uncertainty of system, and the human in torque superposition control also increases the external interferences. Therefore, this paper proposes a LMI robust control algorithm to reduce the external interference and the influence of uncertain factors on the system and improve the tracking performance of system, by use of Lyapunov stability theory and Schur complement property to convert the region pole assignment and robust control constraint conditions into LMI convex optimization problem. The next, the nonlinear vehicle dynamics solving model including Fiala tire model, steering column model is established; the nonlinear tire model is linearized by use of affine function, and the steering superposition control law is solved by use of LMI. Then, the union CarSim and Simulink simulation is conducted under different situations to verify the robustness and control performance of control system. Finally, through establishing the hardware-in-the-loop experiment table based on LabVIEW-RT, the effectiveness of control strategy is verified. The test results show that the method solves the model uncertainty and the robustness decrement problem resulting from human intervention, ensuring a good tracking performance, and a stable system at the same time.