A Compound Diagnosis Method And Its Application In Fault Recognition Of Rolling Bearing

With the abnormal climate,the environmental problem is more and more serious in the world,the demand for new energy is increasing,and the proportion of wind power in the power market is increasing.Wind turbines are mostly installed in areas with harsh environments,which greatly increases the maintenance cost of wind turbines.In order to reduce the maintenance cost of wind turbines,the research on condition monitoring and fault diagnosis of wind turbine is of great significance.In order to ensure the long-term safe and stable operation of wind turbine,the fault diagnosis technology of wind turbine is studied based on vibration signal and pattern recognition.Firstly,a fault diagnosis method based on adaptive variational mode decomposition(AVMD)and Support vector machine(SVM)is studied in frequency domain for rolling bearings of wind turbines.Then,a fault diagnosis method based on Mathematical Morphology(MM)and Correlation Analysis(CA)is studied in time domain.Finally,in order to improve the reliability of fault diagnosis,combining the advantages of the former two methods,a compound diagnosis method based on information fusion is proposed.The main research contents are as follows:(1)A fault diagnosis method of rolling bearing based on AVMD-SVM is proposed according to the non-linear and non-stable characteristics of vibration signals of rotating machinery in wind turbine.This method decomposes the single-component fault signal into multi-component signals with different frequency components by AVMD,then gets the feature vector of the signal by entropy value method,and then uses SVM to train the feature vector to determine the type of signal to be detected.Finally,the classification of signal faults is determined according to the output result of SVM.This method is proved to be effective,but it is not good for the same type of fault with different damage degree(untrained).(2)Considering that the AVMD-SVM method is not effective for the same type of rolling bearing with different damage degree,a fault diagnosis method based on MM-CA is used.The fault signal is processed by MM in time domain,and then the processed signal is transformed into frequency domain to get signal features.Finally,the signal features to be detected and the signal features of each fault class are processed by CA,the fault type with the highest correlation Coefficient is the decision result.The results show that this method has good effect on fault diagnosis of the same type and different damage degree,but its anti-interference ability is weaker than AVMD-SVM method,and its sensitivity to fault is not good.(3)In order to combine the advantages of the above two methods for fault sensitivity and generalization,a compound diagnosis method based on evidence theory is proposed.The improved evidence theory is used to fuse the final diagnosis results of the above two methods to get a new diagnosis result,which improves the fusion accuracy compared with the unimproved evidence theory.The compound diagnosis method combines the advantages of the two methods mentioned above.It has higher recognition rate and reliability than the traditional single fault diagnosis algorithm.