Research On Vibration Characteristic Identification And Fault Diagnosis In Mine Friction Hoist Reducer

Reducer is the key component of mine hoist,its reliability directly affects coal production and personnel safety.The development of on-line monitoring and fault diagnosis system with high identification has an important application value to ensure the safe and efficient operation of coal mine.In this paper,the tower type multi-rope friction hoisting system in the main shaft of Xiaokang Mine of Tiemei group is taken as the research object,and the vibration characteristics,signal processing and feature extraction,fault pattern recognition of reducer are studied in depth.Furthermore,the fault diagnosis technology of mine hoist reducer based on frequency division fusion vibration signal analysis based on combined mode is formed,which provides the theory and technical support for the safe operation of mine hoisting system.The main research contents are as follows:(1)The dynamic model of hoist reducer system was established,considering the influence of external load,gear meshing stiffness and error,fault impact load and vertical vibration of wire rope on the vibration characteristics of reducer under variable speed and load fluctuation.The time-domain and frequency-domain characteristics of the vibration response signals of the reducer working under normal,stationary and local impact fault conditions were obtained.The generation mechanism of meshing modulation band and resonance modulation band was analyzed,and the characteristics of excitation signal and vibration response signal of reducer under normal and typical faults were summarized,which can establish the theoretical basis for subsequent fault feature extraction method.(2)Based on combined mode function(CMF),a fault diagnosis method of hoist reducer based on frequency division and fusion was proposed.The vibration signal was decomposed into two modes of high frequency resonance band and low frequency meshing band by CMF method,and then the white noise added in local mode decomposition(LMD)was optimized to improve the decomposition accuracy.Combined with the multi-scale permutation entropy method,the useful information to reveal the fault of reducer was separated.Based on the fault feature extraction method and the simulation and experimental data,combined with the actual working conditions of the hoist,the vibration signals of different health states in the constant speed and variable speed lifting stages were analyzed,and the accurate positioning of fault characteristic frequency components was realized.(3)In view of the local fault types with different degrees of gear and bearing in hoist reducer system,a compound fault separation method was proposed.The vibration signal was divided into two modulation bands by CMF.Aiming at the indistinct entropy division of the low frequency meshing band arrangement,a denoising and feature extraction method combined with maximum correlation kurtosis deconvolution(MCKD)filtering was proposed to enhance the weak shock features.In view of the high frequency resonance band,the MCKD filter bank preprocessing was proposed to before looking for the best filter band area,and extract the sensitive frequency band location method,so as to realize the effective separation of gear and bearing fault.Through the establishment of reducer fault simulation test-bed for verification,the accurate diagnosis of hoist reducer compound fault was realized.(4)In order to solve the problem of redundant information and high dimension in multi measuring point data of reducer,a method of selecting sensitive feature set based on multi measuring point information was proposed.The method combines distance assessment(DET)with maximum correlation minimum redundancy(mRMR)to eliminate irrelevant features and redundant features.Finally,clustering method was used to classify different faults of the best sensitive feature subset.The experimental data show that this method has better classification effect.(5)Based on the application of on-line monitoring and fault diagnosis technology in Xiaokang Coal Mine of Tiemei group,a real-time monitoring and fault diagnosis system for main shaft hoist reducer was developed.The overall functional framework,signal acquisition,transmission scheme,hardware system and software system of the monitoring system were designed.Based on the Internet of things cloud platform to achieve remote real-time monitoring,the system has been used in the control system of the Tiemei group Xiaokang Coal Mine main shaft hoisting fault to ensures the safe operation of the hoisting system.There are 128 diagrams,11 tables and 153 references in this dissertation.