An experimental and computational study of internal combustion engine modeling for controls oriented research.

摘要

Engine modeling is a long and difficult task with two camps of researchers producing the most commonly used engine models. Original-principles-based models are extensively detailed and require large amounts of computer time to produce results. Alternatively, control-based engine models often rely partly on static experimental data and are computationally inexpensive. However, as engine control schemes become more complicated, it becomes necessary to model more of the fundamental processes in the engine itself. What the controls community needs is a model that is between these two extremes. This dissertation presents a complete engine model that captures most of the dynamics relevant to control engineers. Most importantly, this model is capable of predicting the cylinder pressure based on physical principles. This dissertation presents a two-zone cylindrical flame propagation model that allows for predictive pressure computation and addresses the experimental work required to validate these models. The results show that the simulation underpredicts the engine cylinder pressure during the compression phase, yet overall, the gross IMEP is reasonable.