CRHPC using Volterra models and orthonormal basis functions: an application to CSTR plants

This work focuses on predictive control of nonlinear systems modelled by Volterra series with orthonormal basis functions expansion (Volterra-OBF). By using Volterra series, any nonlinear analytical operator with finite memory can be approximated with an arbitrary precision. A Laguerre basis expansion is then used in the model parameterization to reduce the number of coefficients of the Volterra model. In this context, a predictive control with terminal constraints window based on cited model is proposed. Finally, simulated results using a CSTR unit illustrates the control scheme performance.