Application of wavelet de-noising method in vibration signal analysis of elevator car

During the operation of the elevator, the amplitude of the vibration of the elevator has a direct impact on the degree of comfort when you take the elevator, and the mechanical fault of elevator system will appear through the vibration of the elevator car. So collection and analysis of vibration signal can be used to judge the degree of comfort so as to decide the type of mechanical fault, which can provide the effective evidence for detection of running state and fault diagnosis of elevators. The key point of studying this problem is to de-noise effectively for the vibration signal we have collected; in this way, we can extract fault characteristic information of the signal. From a practical point of view, after collecting vibration signal of a multitude of elevators and conducting a lot of simulation experiments, this paper will demonstrate the wavelet multi-threshold de-noising method works best based on RMS error, SNR and correlation coefficient of honest signal by comparing these three de-noising method: wavelet analysis, wavelet packet analysis and wavelet packet multi-threshold analysis. The practice shows that this method can de-noise effectively and provide reliable signals for detection of running state and fault diagnosis of elevators.