AFR control for motorcycle engine using smith predictor and h∞ control

This paper focuses on feedback control of the air-fuel ratio (AFR) for a motorcycle engine. A feature of this paper is to identify a simple model from the fuel injection time to the AFR, which captures the dynamics in various operating conditions. The model is composed of a single transfer function and a lookup table storing the gain and delay relative to the nominal model as functions of the engine speed, throttle position and water temperature. A compensator for changes of the delay is constructed using Smith predictor. An H_infin controller is designed to handle unmodeled dynamics and disturbances. The effectiveness of the proposed method is investigated by experiments.