LAUSR

Aiming at the problem that the differential term of Newton iterative learning law is not easy to be determined and affects the control performance, the secant iterative learning law is proposed by using the secant method in numerical analysis technology. The proposed control strategy is applied to the speed control of ultrasonic motor to verify its effectiveness. Some practical improvements for the secant iterative learning control strategy are proposed to solve the problems in practical uses of the strategy. According to the nonlinear characteristics of ultrasonic motor, a simple method for determining the value of learning gain is obtained based on the experimental data. Furthermore, online adaptive adjustment method of learning gain is proposed to compensate the nonlinearity of ultrasonic motor. Experimental results show that the motor's control performance progressively improves during the iterative learning process, featuring higher convergence speed and better performance, which indicates that the proposed control strategy is effective.