Polytopic Quasi-LPV Models Based on Neural State-Space Models and Application to Air Charge Control of a SI Engine

This paper is one of two joint papers, each presenting a different representation of a feedforward neural network. Here a discrete-time polytopic quasi linear parameter varying (LPV) model of a nonlinear system based on a neural state-space model is proposed, whereas in the joint paper (Abbas andWerner [2008]) a neural state-space model is transformed into a linear fractional transformation (LFT) representation to obtain a discrete-time quasi-LPV model of the nonlinear system. As a practical application, air charge control of a spark-ignition (SI) engine is used in both papers to illustrate two different synthesis methods for fixed structure low-order discrete-time LPV controllers. In the present paper, the synthesis of a fixed-structure low-order self-scheduledH∞ controller is based on linear matrix inequality (LMIs) and evolutionary search. A controller is designed for the nonlinear system and its performance is compared with that achieved when a standard self-scheduled H∞ controller is used.