Low-Speed Control for Permanent-Magnet DC Torque Motor Using Observer-Based Nonlinear Triple-Step Controller

In speed control of a permanent-magnet dc torque motor, cogging torque is an undesirable disturbance and results in speed ripple. It is especially prominent at lower speeds, with the symptom of jerkiness. This paper provides a novel observer-based nonlinear triple-step controller to improve the low-speed tracking performance. Considering that cogging torque is a fast time-varying disturbance and changes harmonically, a nonlinear parameter-varying high-order system is established to model the fast-varying properties of cogging torque. Then, a reduced-order nonlinear observer is designed to estimate the cogging torque, and the robustness analysis with regard to the uncertainties is given for the proposed nonlinear observer. Thereafter, a triple-step nonlinear method is applied to derive a speed tracking controller, and then the robustness analysis against considered observation errors and lumped uncertainties is completed. Finally, the proposed control scheme is verified through experimental tests, and the results show that tracking errors are substantially reduced at low speeds.