Recognition Methods Of Wearing State For Elevator's Traction Sheaves Based On AE Characteristics

Elevator is an important vertical transportation tool in modern buildings,in recent years,the number of elevator began to increase rapidly with the rise of city high-rise buildings,and the problem of security is increasingly becoming the focus of attention.Traction wheel is not only the core component of elevator system but also an important device to ensure the long-term,safe and stable operation of the elevator system.Once the traction wheel of the elevator is seriously worn down,the traction ability of the elevator system will be greatly reduced and serious threat to elevator operation safety and passenger's personal safety.Therefore,the effective monitoring and accurate identification of the wear degree of elevator traction wheel is an important prerequisite to improve the safety factor of elevator and reduce the accident probability of elevator.This thesis presents a method for monitoring and recognizing the wear degree of traction wheel based on acoustic emission signal and D-S evidence theory.Four groups of acoustic emission sensors are installed on the elevator traction machine and acoustic emission signals of traction wheels with different wear degree were tested and extracted.Wavelet packet decomposition and energy reconstruction method were used to obtain the acoustic emission signal energy characteristics samples of different wear degree traction wheels and the feature sample database were constructed.The membership function model of each acoustic emission signal is constructed with the minimum fuzzy entropy as the optimization objective.According to the fuzzy characteristics of the feature sample,the optimal solution of each membership function is realized.Combined with the D-S evidence theory,the recognition model of the wear degree of the traction wheel is established,and the decision criterion of the system is determined.The acoustic emission signals of four groups of elevators were tested and extracted by field experiments.According to the optimized membership function and D-S evidence theory fusion model,the membership degree,the basic probability assignment and the uncertainty probability of each signal are calculated.Finally,the recognition results are obtained.The experimental results show that,the recognition results obtained by the four sets of experiments are consistent with the actual wear degree.The results show that the recognition model of the tractor traction wheel wear degree has high recognition accuracy,which provides an important technical means for the effective monitoring of the traction wheel and the recognition of the wear degree during the operation of the elevator.