Research On Fault Diagnosis Of Gearbox Of Crane Based On Fault Tree Modeling

With the continuous development of science and technology and the trend of economic globalization,ports have become a link to promote the development of urban economy,regional economy and national economy.The good performance of key components of port large-scale machinery equipment is an important factor to determine the normal operation of port.Gearbox plays a role of power transmission in cranes and is one of the key components of port equipment.The operation status of gearbox directly determines the quality of port machinery and equipment.In gearbox system,the fault factors are complex and diverse,and the probability of compound fault is greater than that of single fault.The following work has been done in this paper:(1)The common faults and failure modes of gearbox are studied,the fault tree of gearbox is established,and the reliability analysis based on the model of gearbox is completed.(2)The selection and extraction methods of fault features are studied to obtain the final sensitive eigenvalues for fault diagnosis.(3)With the help of VB.NET,an intelligent diagnosis expert system for multi-position and multi-size hybrid faults based on fault tree modeling and analysis is designed.(4)The intelligent diagnosis system is divided into two parts: rule reasoning and case matching.Through fault tree modeling and analysis,the reliability of gear box's current state is analyzed and relevant fault information is obtained.Moreover,each node of the fault tree corresponds to several diagnostic rules in the rule base.Combined with K-Means,FCM,SOM and other fault clustering methods,the sensitive features of bearing inner ring,outer ring and rolling body faults and their corresponding diagnostic rules are obtained,and the diagnostic rules based on LMD decomposition are formulated.Then,nine kinds of mixed faults of bearings are diagnosed in detail,including the location,size and degree of faults.A case base of fault cases is established,and the problem cases are stored in the case base with the information of fault features and feature values as symptoms.In order to improve the efficiency of case matching,the case base is classified and retrieved to narrow the scope of matching.In the process of rule diagnosis,eigenvalue decision algorithm is used to obtain the diagnosis vectors,and through the weight of the attributes of each rule and the confidence of the rule itself,the problem of reasoning conflicts that may occur in rule diagnosis can be solved,so as to achieve accurate and fast fault recognition.