Adaptive determination of cut-off frequencies for filtering the in-cylinder pressure in diesel engines combustion analysis

In-cylinder pressure analysis is a key tool for engine research and diagnosis and it has been object of study from the beginning of the internal combustion engines. One of its most useful application is combustion analysis on the basis of the First Law of Thermodynamics. However, heat release law calculations use the in-cylinder pressure derivative signal. Hence, the noise is increased and pressure filtering becomes necessary to remove high frequency noise, thus allowing for accurate combustion analyses. In this work, a methodology to set the cut-off frequency of a low-pass filter is proposed. Statistical criteria are used to separate the signal from the noise through the calculation of the Discrete Fourier Transform of several consecutive in-cylinder pressures cycles. Thus, only physically meaningful information is preserved. The proposed methodology is compared with some adaptive and non-adaptive algorithms used to select the cut-off frequencies, and it shows a good ability to adapt to different engine operating conditions.

►Combustion analysis is performed by means of in-cylinder pressure measurement. ►Filtering is necessary due to high noise and dispersion in the measurement. ►A proposed methodology to select cut-off frequencies is proposed. ►The methodology is compared with other filters, showing a better behaviour.