Novel Generalized Predictive Iterative Learning Speed Controller for Ultrasonic Motors

Abstract

Aiming at the needs of ultrasonic motor's motion control system, generalized predictive iterative learning control (GPILC) strategy is constructed by integrating iterative learning control and generalized predictive control. By designing 2D objective function with the information of the previous control process, it attempts to integrate the generalized predictive control methods such as multi-step predictive and rolling optimization into the iterative learning control law to improve the iterative learning control effect. An effective design method of the iterative learning control law is proposed. Based on 2D prediction model, the differential optimization of the objective function is carried out to derive the GPILC law. Then, based on the nonlinear Hammerstein model of ultrasonic motor, an inverse compensation method for the motor's nonlinearity is designed to realize the effective compensation for motor's nonlinearity. On this basis, a generalized predictive iterative learning speed controller for ultrasonic motors is designed. The results of simulation and experiment indicate that the proposed control strategy and its design method are effective. The iterative learning process of ultrasonic motor's speed response is gradually convergent, and the control performance is good. INDEX TERMS Generalized predictive control, iterative learning control, ultrasonic motor.

Automatic traveling wave excitation of structures with imperfect cyclic symmetry

“…In addition, to overcome performance problems due to dead zone dynamics, which are common in USMs, [14] proposed a nonlinear model of a USM that includes dead-zone dynamics and proposed a controller which dictates both the phase difference and velocity of the motor. An iterative PID controller tuning method was implemented in [15] to overcome unknown nonlinearities in the system. Finally, the effect of the excitation amplitude, phase, and frequency on the speed of rotation of a USM is examined in [16], where a phaselocked loop controller was implemented to track the resonance frequency of the motor.…”

Novel D-SLP Controller Design for Nonlinear Feedback Control

“…[36] proposed a method of torque control for induction machines to compensate prediction error by modelbased prediction. However, the control methods are a complex control algorithm, so they often suffer from high memory capacity [37]- [39] and sensitivity to the initial value, which affects access to the local minima point as a result of system instability [39]. In practice, the memory, convergence and execution time are limited [40].…”