Identification of combustion parameters from cylinder head vibration signal with the influence of random road excitation

Abstract

The combustion process feedback information can be extracted from the vibration signals of the cylinder head. This is of positive significance for the online monitoring of combustion status and closed-loop control of combustion processes in internal combustion engines. To obtain the combustion process feedback information under the random road excitation (RRE) condition, this paper firstly simulates and analyses the influence of RRE on the combustion parameters identification by an established simplified diesel-suspension model. Results show that the amplitude, frequency and phase of road excitation signal seriously affect the identification of the combustion characteristics, and the combustion excitation response signal is covered by the road excitation response signal in most cases. Then, a method of combining a high-pass filter and principal component analysis (HPF-PCA) is proposed to extract the combustion parameters. This method effectively eliminates the road excitation response signal and successfully identifies the start of combustion (SOC), the location of maximum pressure rise rate (LMPRR) and the location of peak pressure (LPP) from the processed vibration acceleration signal. Finally, the phase deviation caused by the filtering is corrected based on the relationship between the phase deviation of SOC and the cut-off frequency of filtering. The corrected results show that the maximum deviation between the SOC identified from the vibration signals that with and without the influence of road excitation is 0.8 deg.