Detection of Inappropriate Working Conditions for the Timing Belt in Internal Combustion Engines Using Vibration Signals and Data Mining

Abstract

Engine damage can result from timing belt cracks, wear, unexpected rupture, and abnormal operating circumstances. In this work, an intelligent system was created to identify and categorize timing belt operating circumstances that involved high loads and high temperatures. For this, vibration Signals were gathered under standard, high load, and high temperature working conditions. Signals from the time domain were transformed into the frequency and time-frequency domains using Fast Fourier. The datamining stage made use of 25 statistical features. were taken out of several signal domains. To choose the optimal features and condense the classifier's input space, the Improved Distance Evaluation (IDE) approach was used. An Artificial Neural Network (ANN) was then input signal features from the time, frequency, and time-frequency domains to assess the efficacy of this proposed approach for identifying the timing belt's faulty operating conditions. The ANN classifier recognized and categorized normal, high load, and high temperature operating conditions with the help of all of these extracted features from time, frequency, and time-frequency domain signals. accuracy of 73.3, 85, and 89.2 percent, respectively. In signals with time, frequency, and time frequency-domain information, the classification accuracy was determined to be 85, 95.8, and 95 percent, respectively. The system's ability to identify and categorize both normal and abnormal timing belt operating situations was demonstrated by the results. A combination of signal processing and feature selection, according to the results, could greatly improve classification accuracy.