Determination Of In-cylinder Pressure Signal Noise With Diffrent Filteration Methods on Internal Combustion Engine

During recording pressure data, for evaluating of combustion process on internal combustion engine, much inevitable noise will be recorded in acquisition data process. Unfortunatly, it will has very bad effects on estimation parameters of predictive thermodynamic models. Thus, It is necessary to pass pressure signal from filteration process before using in predictive models. In this study, we use these four different filterations methods as follwoing: 1- FiltFilt function2- Hilbert filter 3- Three part smothing algorithm4- Five part smothing algorithmThese four methods were carried out on experimental pressure data at four different speeds of engine (TD43) on motoring condition at University of Valenciennes. Then, all methods are evaluated by each other and they are compared base on power of noise cleaning. Forth method, is specially designed for this research, works two times stronger than third method and it is three times stronger than first method too.Numbers of passing signal form filteration process are considered as an important factor. It is specified as 4 turn in this study.To verify pressure signal quality, we use Signal to Noise Ratio. SNR becomes infinity in two remarkable points before and/or after Top Dead Center approximately. This result confirims that maximume in-cylinder pressure signal quality is occured in mentioned points.

Keywords:

In-Cylinder Pressure Signal , Filteration Process , Smothing Algorithm , Noise , Internal Combustion Engine