Cylinder Specific Pressure Predictions for Advanced Dual Fuel Compression Ignition Engines Utilizing a Two-Stage Functional Data Analysis

Abstract

In-cylinder pressure is a critical metric that is used to characterize the combustion process of engines. While this variable is measured on many laboratory test beds, in-cylinder pressure transducers are not common on production engines. As such, accurate methods of predicting the cylinder pressure have been developed both for modeling and control efforts. This work examines a cylinder-specific pressure model for a dual fuel compression ignition engine. This model links the key engine input variables to the critical engine outputs including indicated mean effective pressure (IMEP) and peak pressure. To identify the specific impact of each operating parameter on the pressure trace, a surrogate model was produced based on a functional Gaussian process (GP) regression approach. The pressure trace is modeled as a function of the operating parameters, and a two-stage estimation procedure is introduced to overcome various computational challenges. This modeling method is compared to a commercial dual fuel combustion model and shown to be more accurate and less computationally intensive.