A combined feedforward-feedback controller design for nonlinear systems

Abstract

This study presents a control framework that integrates both feedforward (FF) and feedback (FB) procedures to control nonlinear systems. The intelligent control component consisted of a modified recurrent wavelet neural network used to build the FF controller. Furthermore, the H-infinity control, renowned for its robust and durable attributes, was employed in the FB loop. The parameters of the controller were improved using particle swarm optimization. The proposed controller’s efficacy was compared with that of another FF-FB controller, in which the FF controller was a multilayer perceptron, and the FB controller was an H-infinity controller. The effectiveness of the proposed framework to control the nonlinear mass–spring dynamics was evaluated by analyzing its control accuracy and its capacity to tolerate external disturbances.

Keywords

• pso
• robust controller
• intelligent controller
• wavelet neural network
• h-infinity