Artificial neural network for piezoelectric control systems

This paper presents a neural network controller for a piezoelectric controlled structure by emulating the control performance of a Linear Quadratic Gaussian (LQG) controller. The configuration of the Artificial Neural Network (ANN) is simple, yet it is efficient in terms of its high learning speed and good generalization ability. A case study is presented to demonstrate the performance of the ANN controller versus the LQG controller. The test results for different disturbances on the structure show excellent agreement between the ANN and LQG controllers.