Base On Mathematics Appropriate Fitting Method Gas Turbine Blade Failures Diagnosis

Gas turbine is one of the power plant in the field of aviation and ship,its status andimportance have been payed more and more attention, gas turbine technology has been rapiddevelopment in recent years. In the gas turbine field, some western countries started relativelyearly, A lot of achievements have been maken during the development. Our country gasturbine research base is relatively weak, but through continuous exploration also made a lot ofresearch and theory.Research the gas turbine structure and working methods, can be found the gas turbineblades is a major component in the gas turbine. High-temperature gas is depended the leavesturn heat into mechanical energy. And because the gas turbine irreplaceable position in thecivil and military fields, how to improve the output power and efficiency of the gas turbinehas become a major research and development trends. The most direct way to improve gasturbine performance is increasing the temperature and pressure of the gas, however, limitedby the gas turbine component production materials, especially,gas turbine blade have to worklong hours under high temperature and high pressure environment is very prone to failure,once the leaves are abnormal, will directly affect the stable operation of the gas turbine.Therefore, monitoring and diagnosis on the state of the gas turbine blades is an importantguarantee of the gas turbine in the steady state. Assess accuratelt the fault information of theblade failures have great significance.The main topic is using the theory of fault diagnosis in the detection for the temperatureof gas turbine blades, select the method of fault diagnosis based on the analytical model, inthe gas turbine working properly, establish the analytical model leaf temperature distributioncurve, then compare the actual collection of temperature parameters with the model, with thecomparing process to achieve the monitoring of gas turbine blade temperature state, and thetemperature abnormal position of judge. Process to establish the temperature distributionmodel, using curves of the vector fitting method, build on the temperature formula. Use fittingsoftware1stOpt, it is a mathematical optimization analysis tool package, widely used in thefield of curve fitting, do not need to set the initial value in the fitting process will be able toget the results, greatly improves the fitting accuracy. During the process of data fitting, analyzed the leaf sample temperature characteristics and distribution trends, throughcomparing groups of sampled data, finding in different conditions and different rotationperiod each other’s differences, split leaves temperature data, extract the temperaturecharacteristics of a single leaf, in order to constantly improve the mathematical model of leaftemperature curve. At the same time compare the collected failure data with the curve ofmodel, comparative analysis on the judgment of the leaf temperature monitoring and faultlocation.