Research On Fault Diagnosis And Prediction Method Of Gear Box For Large - Scale Hoisting Equipment

Gearbox of large crane’s lifting mechanism frequently occurs to failures while the structure of gearbox is complex and the crane is often under harsh working conditions. Traditional diagnostic methods for gearbox are inefficient, limited in diagnostic range, and unable to detect incipient failures. Therefore, research on efficient and reliable methods of fault diagnosis and prediction for large crane’s gearbox system is of great significance.Firstly, taking the gearbox of large crane as research object, common faults and causes are analysised, then an overall program designed for large cranes’s gearbox fault diagnosis and prediction methods is carried out. Secondly, as the gearbox temperature data and oil data are nonlinear and non-stationary, an OSVR(Online Support Vector Regression) prediction method based on combined kernel functions is proposed using different kernel characteristics, and the method is applied for predction of temperature and oil status. Thirdly, A separation method which combines EMD(Empirical Mode Decomposition) and optimized ICA(Independent Component Correlation Algorithm) is proposed to solve the separation problem of fault composition and noise from vibration signals, and then spectral analysis and envelope analysis are used for fault feature extraction. Fourthly, for the intelligent diagnostic classification of gearbox’s vibration data, the construction of feature vectors based on energy entropy of spectrum has been reseached, and a fault diagnosis method based on higher-order cumulant spectrum and SVM classifier is proposed.The experimental results show that the method has good diagnostic performance. Finally, a fault diagnosis system for gearbox of large crane is developed, which integrates the data-driven fault diagnosis and prediction methods.The reliability of the methods and the applicability of the system are verified through the practical example from work site, which not only provides technical reference for safety monitoring of large crane’s gearbox, but also possesses great engineering application value and practical significance.