عنوان مقاله [English]
نویسندگان [English]چکیده [English]
In this research, an intelligent and automatic procedure is introduced for diagnosis of some common faults of timing belt bases on its vibration signals. For this goal, vibration signals were gained in different faulty conditions of timing belt. In data mining step, six features namely, standard deviation, kurtosis, skewness, impulse factor, shape factor and crest factor were extracted from vibration signals aim to better monitoring of signals behavior. After extracting the fault characteristics, Artificial Neural Network (ANN) classifier was used for intelligent detection of faulty belts. The ANN was trained with 60 percent of signals and was tested with other signals. The results show the faulty belts have a turbulence vibration and impulsive behavior. Also, ANN classifier detected and classified the timing belts faults with 90 percent accuracy. The results show which the use of vibration signals for small faults detection of timing belt was effective aim to prevent its complete rupture. The results demonstrate the combination of data mining and artificial intelligence techniques is a powerful and successful procedure for precision timing belt troubleshooting.