Dynamic Test And Fault Forecast Analysis Based On Data Mining Technology

Data mining technology is a research aspect, which appears in database fields recently, and it has become a hotspot field which the internal or overseas scholars are researching into. The author of this thesis ameliorated the existing data mining arithmetic on the basis of researching data mining arithmetics, and had application in fault problems of equipments on chemical technology. The main aim of the thesis is ameliorating the existing arithmetics, and has application in equipments running detection of the enterprise of chemical technology to avoid economic lost from the fault of equipments.This thesis is mainly focus on the fault forecast problem of compressor, that is, on the basis of the relationship of the gathering correlative parameters of compressor (such as compressive stress of exit or entrance,temperature,oscillation of axletree,the quantity of leaking gas and so on), concluding the possible reason of the fault, taking effective measures to avoid economic lost from shutting down of the fault of equipments. By researching the data mining arithmetics, analyzing and comparing, the thesis filtrated a rather appropriate arithmetic------Decision Tree, and ameliorated the arithmetic rationally to have an effective application in this system.At present, the system of fault diagnoses and forecast analysis of compressor has been opening out and is having application in the system of equipments running detection and manage of the high pressure workshop of plastic factory of SINOPEC QiLu company, and in the actual producing period it will avoid the fault of compressor effectively.