Adaptive internal model control design for positioning control of a piezo-ceramic actuator with rate-dependent hysteresis

Abstract

This paper proposes the novel design of a positioning control system of a piezo-ceramic actuator. Piezo-ceramic actuators have been used extensively in many engineering applications which require high precision positioning. However, it is well known that piezo-ceramic actuators exhibit hysteresis in their response to driving input which considerably deteriorates positioning accuracy if no appropriate compensation has been made. Huge efforts have been devoted to the design of control system for positioning control of a piezo-ceramic actuator and a number of different control strategies are proposed for its hysteresis compensation. This paper tackles the problem of the compensation of rate-dependent hysteresis of a piezo-ceramic actuator. Rate-dependence of a hysteresis means that its behavior will vary if the rate or the frequency of the driving signal of the actuator changes. This paper proposes the use of two radial basis function neural networks (RBFNN) to construct a high precision positioning control system of a piezo-ceramic actuator which exhibits rate-dependent hysteresis. The proposed control system takes the form of the internal model control (IMC) system, where one RBFNN is used as the internal model of a piezo-ceramic actuator having rate-dependent hysteresis while the other RBFNN is configured to work as a controller of the system. It is shown in this paper that RBFNN model trained with particle swarm optimization (PSO) algorithm provides good modeling performance for a wide range of driving frequency, and the RBFNN controller is made adaptive with back propagation on-line parameter updater to cope with possible modeling inaccuracy between the actuator and its model. Results of the positioning control experiments indicate that proposed adaptive internal model control system with two RBFNNs shows adequate performance on the compensation of rate-dependent hysteresis to guarantee high precision positioning.

Keywords

• adaptive control
• rate-dependent hysteresis
• piezo-ceramic actuator
• rbfnn
• pso
• imc