A Tire Tread Production Fault Diagnosis Expert System On SVM

The diagnose technology of equipment state has big development at home and broad in recent years and has brought big roles into play in the aspect of ensuring that equipment safety,stable,longtime high grade work.With technology progress and development,machinery equipments tends to be large-scale-rization,automation, hign-speeding and complicate,This makes equipment fault diagnosis technology become more and more important.In the background of PLC controlling system of Tire Extruding Line and according to system population structure and function characteristic,this paper analyses the types,range and causes of the faults in the line.On the basis of now having all equipment diagnosis theory merit and deficiency,we concentrate on introducing the expert system and the theory of SVM(support vector machine). Connecting with study background,we established the fault diagnosis model of the system by using expert system and SVM.We have summarized the creation and present situation of intelligence fault diagnosis system,analysed the general structure of fault diagnosis expert system,and We have been completed the population structure and function design of expert fault diagnosis system,including the Knowledgebase,Database,Knowledge acquisition, Reasoning machine and etc.In this paper,we have applied SVM theory into fault digonsis system of Tire Extruding Line and set up the model of Multiple classification.We discussed the characteristic of two kinds of multiple classification,including one-vs-one classificati -on algorithm and one-vs-rest classification algorithm,and then wo given the step of two kinds classification algorithm.We are dedicated in bolt extrusion machine fault dignosis,analysed operating principle and the types of extrusion machine in the work. We Have collected datas of extrusion machine from the scene:include on Rotation rate,pressure,electric current,and have simulation on the data through Matlab prog -ramming software and SVM Toolbox.And then we analysed the classification results and the function of one-vs-one classification algorithm and one-vs-rest classifycation algorithm.Two kinds classification algorithm all are able to reach very good anticipating effect through adjusting the parameter of kenel function.The fault diagnosis cases of extrusion machine of tire tread are indicated that the two sorts of multi-variety algorithm of SVM have quite different diagnosis results in the same samples.Besides,with application of small-sample fault diagnosis,SVM is more adaptive and holds better generalization capability.