In-Cylinder Pressure Based Engine Knock Classification Model for High-Compression Ratio, Automotive Spark-Ignition Engines Using Various Signal Decomposition Methods

Engine knock determination has been conducted in various ways for spark timing calibration. In the present study, a classification model was developed using machine learning algorithm. Wavelet packet decomposition (WPD) and ensemble empirical mode (EEMD) were employed characterization of in-cylinder pressure signals from experimental engine. The WPD used to calculate 255 features seven levels. EEMD provided total 70 their intrinsic functions (IMF). engine operated at advanced timings induce knocking under speeds load conditions. Three intensity metrics determine that dataset included 4158 cycles out 66,000 cycles. trained with achieved an accuracy 99.26% cycle detection. neighborhood component analysis revealed contributed significantly classification. retrained significant 99.02%. Although misclassification rate increased normal detection, feature selection decreased size 253 8.25 MB. Finally, compact 99.95% second obtained borderline (KBL) timings, which validates is sufficient KBL determination.