Research On Fault Diagnosis Of CNC Machine Tools Based On Performance Characteristics And Fault Symptoms

With the rapid development of modern manufacturing industry, CNC machine toolshad been widely used. Due to the complexity of the structure and function of the CNCmachine tools, the diagnosing work of the failure will be difficult. So to diagnose the faultafter the occurrence of the events is a major problem faced by the manufacturing process.Considering the complicated process, this paper carried out a series of research work.First, this article analyzed the characteristics of CNC machine tools and elaborated thewidely discussed diagnosis methods of CNC machine tools. After comparing the differentcharacteristics of each method, expect system (ES) based on fault symptoms andintelligent diagnostic algorithm based on performance characteristics value have beenapplied in the research. Secondly, the relationship between performance characteristicsand faults was discussed. SOM-BP artificial neural network was built up and trained,realizing the function of fault diagnosing of NC machine tools. And accuracy was taken asthe example to verify the method. Thirdly, the relationship between the fault symptomsand fault was expounded. Expect system combined RBR and CBR technology was builtup to realize the fast-diagnosing function and the sustainable growth of knowledge basebased on fault symptoms of CNC machine tools. And then the article described thedeveloping of “Heavy CNC machine fault diagnosis system”. Except for the diagnosisfunction, this software also realized the management module of CNC structures, thefailures and the diagnosis rules, and was practical in manufacturing process. At lastsummary and outlook was presented.