Model-based calibration for the control variables of an Atkinson cycle engine

Abstract

An Atkinson cycle engine has been designed for use in hybrid vehicles. The Atkinson cycle engine is generally operated at part load operating conditions. A novel load control strategy that combines the intake valve operation and electrically throttling control has been investigated for the purpose of improving the fuel economy at part loads. Other control variables such as spark angle also significantly impact the fuel economy. These control variables highly correlate with each other, if accurate calibrations of these control variables are conducted only through practical experiments, the costs will largely increase. This paper proposes a physical model and genetic algorithm-based automatic calibration methodology. The physical model was accurately calibrated at a series of speed-load points to match experimental results. Then the fuel economy was further optimised based on the physical model. After the optimisation, the experimental fuel economy is obviously improved with the maximum as 7.67%.