| The traditional fault diagnosis method not only consumes manpower and material resources,but also leads to serious waste of resources,and the fault diagnosis rate of the traditional fault diagnosis method is relatively low.In recent years,intelligent fault diagnosis algorithm has become one of the research directions for fault diagnosis of complex systems.The component composition and circuit logic of the automatic loading system of armored equipment are complex.Considering the result of fault diagnosis,it is necessary to give suggestions for component-level maintenance,so as to accurately locate the fault and achieve the purpose of state monitoring and fault diagnosis of the automatic loading system of armored equipment in a relatively comprehensive way.Firstly,through the detailed analysis and comparison of common state monitoring and fault diagnosis algorithms,the shortcomings of each method in practical engineering application are proposed.Based on the actual control characteristics of the research object,the feasibility of the proposed state monitoring and fault diagnosis algorithm is analyzed and demonstrated.Secondly,based on the structure and control principle of automatic loading system was analyzed,the failure mode in detail,puts forward the attributes reduction of rough sets to eliminate redundant knowledge acquisition method of attribute reduction on the knowledge obtained for effective interval to integrate and optimize the initial cut sets,discrete reduction is obtained after the extraction of rules.The fault tree analysis method is used to analyze the causality of parts and find out the direct cause of the fault.The goal of improving fault diagnosis speed and decision accuracy is achieved.Thirdly,an improved Jeffery evidence updating rule algorithm with time tag is proposed to ensure the real-time performance of fault diagnosis and the authenticity and credibility of evidence.The current evidence obtained by fault diagnosis is used to update the fault diagnosis evidence obtained at the last moment in real time,and the basic confidence of updated diagnostic evidence is converted into Pignistic probability for fault decision-making.The algorithm improves the defects of general fuzzy theory and traditional evidence theory,and the obtained fault diagnosis results are accurate,real-time and reliable.Finally,a complete automatic loading status monitoring and fault diagnosis system of armored equipment is formed.The system uses Visual Studio 2010 software and SQL Server 2008 database for development.The system has been tested and implemented effectively,with the characteristics of flexible operation,simple and convenient.In this system,the dynamic fault diagnosis algorithm based on improved evidence updating rules is applied to solve the problem that the causes of the faults of armored equipment loading system are complex,and the single factor and single model are not enough to accurately locate the component-level faults.It is not hard to see from the comparative analysis of the fault diagnosis results that the fault diagnosis algorithm proposed in this paper has obvious advantages and plays a role in solving the dynamic fault diagnosis problem. |
