| In this paper, a novel control method is introduced to design controllers for reduced cost and improved performance and robustness of electrical power assist steering systems. The method employs estimators to generate accurate estimates of driver's torque input and eliminate the need of pinion torque sensors. Together with measured signals such as steering wheel rotational position and/or electric motor velocity, these estimates provide signals for a robust controller that consists of a feedforward part vs. the boost curve and an H∞ feedback to ensure refined performance, robustness, and disturbance attenuation.; Introduction of estimators results in additional feedback loops and ushers in new issues of robust stability and performance. The concept of structural stability is introduced to capture the requirements of robust stability under various system configurations due to component failures. The requirements will guarantee a level of fault tolerance in case of control failure during its operation. Conditions are provided under which such structural stability is maintained.; Since the torsion stiffness and the moment of inertia changes as a function of steering angle, a Least Square Algorithm method was used to estimate these parameters. Finally, to achieve the desired steering feel as in the conventional hydraulic steering system, a gain scheduling method was developed to capture the precision feel of “on-center” and “off-center” steering feel. |
