Robust H∞ Control for the Turbocharged Air System Using the Multiple Model Approach

This work deals with the modeling and control of the air system of a turbocharged spark ignited (TCSI) engine. The main purpose is to design a set of nonlinear controllers based on the original physical models of the systems. The proposed strategy is based on the multiple model control approach. Linear Matrix Inequality (LMI) conditions are derived from the Lyapunov's direct method in order to compute the controller for the multiple model subject to bounded noises. The disturbances attenuation is addressed via an H-infinity-constraint. In comparison with existing approaches, the proposed strategy handles more easily the nonlinearities and alleviates significantly the calibration and implementation tasks. Moreover, the feedback gains are readily tuned with few parameters. Finally, this method can be generalized to others more complex turbocharging systems with some adaptations.