Nonlinear adaptive internal model control using neural networks

Abstract

The IMC structure, where the controller implementation includes an explicit model of the plant, has been shown to be very effective for the control of the stable plants typically encountered in process control. A nonlinear internal model control(NIMC) strategy based on neural network models is presented for SISO processes. The nonlinearities of the dynamical system are modelled by neural network architectures. Recurrent neural networks can be used for both the identification and control of nonlinear systems. Identification schemes based on neural network models are developed using two different techniques, namely, the Lyapunov synthesis approach and the gradient method. Both identification schemes are shown to guarantee stability, even in the presence of modelling errors. The NIMC controller consists of a model inverse controller and a robust filter with single adjustable parameter. Using the theoretical results, we show how an inverse controller can be produced from a neural network model of the plant,without the need to train an additional network to perform the inverse control. This NIMC approach is currently restricted to processes with stable inverses and with relative degree equal to one. Computer simulations demonstrate the proposed design procedure.