Performance Evaluation And Fault Diagnosis Of Gear Box Based On FCE And DUCG

Gear box is used to transfer power and movement,as well as the most widely used driving part,which plays an important role in the field of industrial production and engineering application.If there is any fault in the gear box,it could lead to the whole equipment doesn't work,even make the significant safety accident happen.The study of running state's evaluation and fault diagnosis of gear box is beneficial to improve the running security,reliability and stability.By the method of fuzzy comprehensive evaluation,the gear box which equipped with normal,wear and tear's gear had been comprehensive evaluated,and Dynamic uncertain causality graph had used to fault diagnosis for gear box in this paper.The main research contents are as following:(1)The Hidden Markov Model is applied to evaluate the state of the operational gear box,using piezoelectric acceleration sensors to get the vibration signal of the gear box under different state,taking the algorithm of Baum-Welch to adjust the normal,wear and tear's parameters after get the observed value and the initial conditions.Forward-Backward Algorithm is used to get the similarity among different running state,by comparing the degree of similarity between different running state to verify the rationality of the quantitative results.Converting the similarity to the density of infection and the result had been converted to the degree badness,adopting the combined membership functions of half trapezoid and middle ridge shape distribution as the membership degree in different membership grades.Taking the comprehensive close degree to make a final assessment of the state to the target,and the situation that the maximum membership principle can't be applied in extreme cases is avoided.Experiments show that the tear's state is more serious than others,and the comprehensive close degree is more sensitive to identify different running state.(2)The Dynamic uncertain causality graph is applied to fault diagnosis for gear box,considering the coupling relationship of the different faults and the complex structure of gear box,choose the gear,bearing,shaft and box to analysis,and their own sub-DUCG had been set on.Analyzing the complex logic relationship between gear,bearing,shaft and box,by using logic gate variables and condition connective variables,all the separately sub-DUCGs had been constructed to a whole DUCG.Presently,the problem that the correlation between variables is lack of objectivity is focused,considering the necessity of the combination of qualitative and quantitative analysis,the new DUCG is proposed based on the method of combination weighting which is verified by specific examples,and the different weighted number of fault had been chosen to analyze,the limitations of the diagnosis accuracy and the change of the weight had also been discussed,research shows that the probability growth and growth rate of the biggest primary event increases gradually along with the increase of the number of weight in addition to the special circumstances,which provides a higher accuracy method for the fault diagnosis of gear box.