Search Results

This work focuses on the development and validation of a data-driven model capable of predicting the maximum in-cylinder pressure during the operation of an internal combustion engine, with the least possible computational effort. The model is based on two parameters, one that represents engine load and another one the combustion phase. Experimental data from four different gasoline engines, two turbocharged Gasoline Direct Injection Spark Ignition, a Naturally Aspirated SI and a Gasoline Compression Ignition engine, was used to calibrate and validate the model. Some of these engines were equipped with technologies such as Low-Pressure Exhaust Gas Recirculation and Water Injection or a compression ignition type of combustion in the case of the GCI engine. A vast amount of engine points were explored in order to cover as much as possible of the operating range when considering automotive applications and thus confirming the broad validity of the model.

During the development of an Internal Combustion Engine-based powertrain, traditional procedures for control strategies calibration and validation produce huge amount of data, that can be used to develop innovative data-driven applications, such as emission virtual sensing. One of the main criticalities is related to the data quality, that cannot be easily assessed for such a big amount of data. This work focuses on an emission modeling activity, using an enhanced Light Gradient Boosting Regressor and a dedicated data pre-processing pipeline to improve data quality. First thing, a software tool is developed to access a database containing data coming from emissions tests. The tool performs a data cleaning procedure to exclude corrupted data or invalid parts of the test. Moreover, it automatically tunes model hyperparameters, it chooses the best set of features, and it validates the procedure by comparing the estimation and the experimental measurement.