Fuzzy linear quadratic tracking control based on an extended state observer for automotive electric power steering

An electric power steering (EPS) system provides superior performance compared to other traditional steering systems. Despite their wide use, controllers such as proportional–integral–derivative (PID), linear quadratic regulator (LQR), sliding mode control (SMC), and active disturbance rejection control (ADRC) still exhibit drawbacks such as phase lag, chattering, and high sensitivity to noise, especially when applied to multi-input multi-output (MIMO) systems. This study introduces a novel approach that integrates fuzzy logic with linear quadratic tracking (FLQT) and employs an extended state observer (ESO) for state estimation to address the aforementioned challenges. The proposed control framework effectively suppresses transient errors, eliminates phase delay, and enhances robustness against external disturbances and sensor noise. The novelty of the proposed control method lies in its ability to simultaneously handle multiple non-linearities and imperfections while ensuring the system stability. Simulation results demonstrate that the output signals (state variables) closely track the desired signal with negligible error with the application of the proposed control approach even in the presence of external disturbances. The root mean square (RMS) tracking error of the steering column and steering motor angles does not exceed 0.2 % at low vehicle speeds and remains below 0.5 % at high vehicle speeds. In addition, the error between the estimated signal (state variables and total disturbances) and the actual value is less than 5 %. The variations in assisted torque achieved from the FLQT-ESO based controller follows the pre-set reference value. The system stability is guaranteed when inputs (vehicle speed and driver torque) change. In conclusion, the proposed approach is highly effective in controlling the automotive steering system and can be applied to several other mechatronic systems.