Non-Intrusive Indirect Adaptive Control of an Uncertain Gasoline Engine Three-Way Catalyst

In this paper we develop an indirect adaptive control architecture for gasoline engine emission control. A model-based PID feedback controller is presented where its calibration amounts to finding the parameters of a simple catalyst model consisting of a gain and time delay, followed by an integrator. Both the gain and time delay depend nonlinearly on the engine air mass-flow rate, and therefore, are parameterized by lookup tables for efficient online implementation. We develop an indirect adaptive controller to account for uncertainty in the parameters as the catalyst ages. In order to maximize catalyst conversion efficiency, a small-amplitude square-wave input is imposed on the front face of the catalyst; the estimation routine takes advantage of this signal to non-intrusively identify the catalyst gain lookup table online, which is subsequently used to autotune the PID control gains. Simulation results with realistic vehicle inputs are demonstrated with parametric model uncertainty.