Skip to Main Content
Aiming at the fault features of the elevator braking system, the basic characteristics of three faults types are analysised. By detecting the brake shoe gap-time signals in the process of braking, the fault signals are decomposed using wavelet packet, and the signal characteristics of 8 frequency components from the low-frequency to high-frequency in the third layer are extracted. Then taking advantages of B-spline and fuzzy neural networks to set up the elevator braking system fault diagnosis model, the 8 obtained eigenvalue are used as the model inputs for fault diagnosis. The result shows that this method is effectual and applied.