PID Controllers Using Neural Network For Multi-Input Multi-Output Magnetic Levitation System
AbstractThis paper deals with a design schemefor PID controllers with neural network (NN) for multi-input multi-output (MIMO) magnetic levitation systems. PID controllers are considerably used in industrial processes because their simple structure that consists of only three parameters. So far, considerable attention has been given to the use of SISO procedures for the tuning of decentralized PID controllers for MIMO systems. This approach suffers from the need for full model knowledge of the plant to compute the ultimatepoint, and it finally uses only this information to design the controller. Here, we develop a controller without requirements of the full model knowledge of the plant. We develop a NN that can be used to assist PID controllers for MIMO systems. The NN is used to compensate for inputs-outputs coupling of the MIMO system and to stabilize the PID controllers. Finally, the effectiveness of the proposed controller is confirm through experimental studies.