LPV controller synthesis for charge control of a car engine - A hybrid evolutionary algebraic approach

This paper presents the design of an LPV controller for the intake manifold of a spark ignition engine. A nonlinear model of the plant is - with slight simplifications - turned into an affine LPV models in the form of an LFT representation. Using a hybrid evolutionary-algebraic synthesis approach that combines LMI techniques based on K-S iteration with evolutionary search, a scheduled PID controller is designed. A systematic method for initializing the search is discussed, that generates a set of LTI controllers which is then approximated in a least squares sense by an LFT controller. Nonlinear simulation indicates that the scheduled PID controller achieves the performance requirements.