An Efficient Linearized Volterra Series for Nonlinear Model Predictive Control in Dynamic Systems

Abstract

Non-Linear control is a significant part of every industry’s control process. A number of methods have been introduced to model the nonlinearity existed in the control model. Predictive control which is a model-based control method is one of the advanced control methods in practical applications because it can be implemented in time-domain, and easy constraint handling during the design process, as well as the capability of implementing multi-variable systems. A linearized Volterra model for NMC can be proposed to convert the nonlinearity of the system into a series of linearly equivalent expressions. Obtained results show that linearized Volterra series can effectively model dynamic non-linear systems which further allowed employing the proposed model for predictive control. The results reiterate that the use of the Volterra series is comparatively simple and promising in terms of non-stable system modeling.