Study On Expert PID And Iterative Learning Speed Control Of Ultrasonic Motor

As a new type of micro motor, ultrasonic motor has been recognized and appliedin many aspects such as the aerospace, the intelligent robot, etc. The process ofelectromechanical energy conversion of ultrasonic motor is complex; it has obviousnonlinearity and time-varying characteristic. To improve its control performance, somecomplex control strategies have been proposed, such as the neural network. But thecalculation of these methods are complex, this lead to the increment of system’shardware cost and it goes against the industrialization of ultrasonic motor. In the studyof motor control, the control strategy is always expected can not only meet to theexpected performance, but also have little computation to reduce the complexity of thesystem, thus the cost can be reduced and the reliability of the system can be improved.In this paper in order to obtain a control strategy with little computation and highcontrol performance, on the basis of the study of the performance of the fixed PIDcontroller and through the theoretic analysis of the typical control process and thestudy on the speed control experiments of ultrasonic motor, a simple expert PID speedcontroller with two expert rules of ultrasonic motor was proposed. According to thecurrent control conditions, the parameters of the PID control are adjusted by the expertrules and the control performance is improved. Compared with the traditional PIDcontroller, the online computational amount of the proposed method has a littleincrement and it is easy to be realized by devices such as SCM. Experiments undervarying working conditions show that, the proposed control method can well deal withthe change of property and the load disturbance of ultrasonic motor, the control effectis good.The iterative learning control method of ultrasonic motor is studied further. Thealgorithm of the iterative learning control is simple and the values of correspondingparameters can be determined on the basis of little experimental knowledge, so that the complex controlled object such as ultrasonic motor can be controlled effectively.Considering the control characteristics of ultrasonic motor, three rotational speedcontrol laws are designed, they are the P-iterative learning, the PI-iterative learningand the nonlinear iterative learning. Experiments show that because of the essentiallyopen-loop control, the simple nonlinear iterative learning law may bring about thesteady state error. The P-iterative learning law and the PI-iterative learning law canmake motor’s system have good ability to learn, the control effect is improvedgenerally in the process of study and the algorithm is simple and easy to be realized.