An effective data analysis technique is presented in this paper for condition monitoring (CM) and fault diagnosis of a quayside container crane (QCC) by utilizing a combined empirical mode decomposition and wavelet packet quantization. The technique is used to extract useful features from the real-time, nonlinear vibration and temperature data from a lifting motor of a crane to improve the reliability in quayside container operation. It is found that the multiple correlation coefficient (MCC) extracted from the real-time CM data of the lifting motor using the proposed technique can be utilized for an accurate fault diagnosis of the motor. It is also shown that when the QCC motor is at healthy state, the MCC calculated from the vibration and temperature data is close to unity and the variation trends (i.e., the change of the signal energy) calculated from these CM data are similar. On the contrary, the variation trends of these CM data become different from each other when the motor has a fault and the MCC value calculated from these data is far less than unity. The study also shows that the feature extracted using wavelet packet quantization technique alone can lead to erroneous results.
© 2001-2024 Fundación Dialnet · Todos los derechos reservados