Research On Performance Degradation Of Gas Turbine And Trend Forecasting Technology

Due to the harsh and varied operating environment, the components of gas turbine will degrade with a period of operation, and this has a bad impact on overall performance, not only the economy but also the reliability and safety of the gas turbine will decline. Therefore,the research on gas turbine performance degradation and trend forecasting technology has a very important practical significance.The paper will study the degradation of gas path components of gas turbine and the associated trend forecasting technology, the main work is as follows:(1) The research on gas path component degradation of gas turbine modeling method.The performance parameters of gas path components will change with gas path component degrading, so that the overall performance of gas turbine will degrade and measurable parameters of gas turbines will also change. Therefore, we need build a gas turbine gas path degradation model to get the relational model between changing trend of the gas path performance parameters and changing trend of measured parameters. First of all, we should deal with thermodynamic formulas of all gas path components using small deviation linearization method. Then we should isolate independent performance parameters when performance degradation occurs. Finally, this paper establishes the gas turbine mathematical model of the deviation of measurement parameters and the deviation of performance parameters combined with a series of balance relations (the absolute flow balance, power balance, balance, pressure ratio balance and so on).(2) The research on gas turbine performance simulation because of compressor performance degradation. The compressor fouling is the most common reason that causing performance degradation of gas turbine. The compressor fouling is a soft degradation,has a continuous degrading track. Therefore, there are the feasibility to predict fouling degree. The mechanism of gas turbine engine compressor fouling, fouling sources, the sensitivity of compressor and the influence of fouling on the performance of gas turbine components are analyzed and summarized in this paper. Finally, a compressor fouling performance degradation model is established, and the measurement parameters' degradation regulation of gas turbine during compressor fouling is obtained through simulation.(3)The research on single-parameter performance degradation prediction based on time series model. The paper studies the single-parameter time series model method's application of trend forecast of exhaust gas temperature, fuel flow, low pressure compressor efficiency and low pressure compressor pressure ratio. The paper utilizes Eviews 6.0 software to establish the e exhaust gas temperature, fuel flow, low pressure compressor efficiency and low pressure compressor pressure ratio time series ARIMA model and achieves the forecast of future gas turbine exhaust gas temperature, fuel flow, low pressure compressor efficiency and low pressure compressor pressure ratio trend. Then the gas turbine cleaning time is obtain.(4) The research on multi-parameter performance degradation prediction based on SVM. The paper studies the multi-parameter and single-parameter SVM method's application of real-time trend forecast of exhaust gas temperature, fuel flow, low pressure compressor efficiency and low pressure compressor pressure ratio. First of all, a comprehensive forecasting method based on multi-parameter and single parameter is obtained through different single-parameter and multi-parameter EGT predictions on different data sample.Then we can obtain whether the next time is the gas turbine cleaning time based on comprehensive forecasting method.