Research On Side Pancel Crack Fault Diagnosis System Of Large Scale Vibrating Screen DZK2466

This dissertation systematically studies the side panel crack fault diagnosis of large scale vibrating screen. It takes the large scale linear vibrating screen DZK2466of Tianzhuang Coal Preparation Plant of Pingdingshan Coal Mining Group Company as research object. The side panel cracks are divided into four levels by requirements of field actual production and equipment maintenance. The side panel vibrating data at different crack level state has been obtained by field tracking and detection. The side panel crack fault diagnosis method with small samples of vibrating screen based on AR model, PCA and SVM theory is proposed. The author studied wavelet and wavelet packet analysis theory and implemented wavelet de-noising and wavelet packet energy feature extraction. Neural network and genetic algrithoms are studied and the wavelet genetic neural network is established which realizes quick and accurate diagnosis of fault. Finally, the dissertation proposes an online intelligent fault diagnosis method of vibrating screen based on database. This kind of system processes, calls, stores and analyzes the online information using database, VB and Matlab software with friendly interfaces and easy operation. It implements the online diagnosis and accurate early warning to side panel of DZK2466, especially the accurate early warning of fatigue crack has very positive significance to safety running of vibrating screen equipment.