基于时间序列模型的燃气轮机气路性能退化预测

Prediction on Gas path Performance Degradation of Gas Turbine Based on Time Series Model

针对燃气轮机实际运行过程中压气机积垢导致的性能退化与定期清洗维护问题,提出一种基于时间序列模型的燃气轮机气路性能退化预测方法。并在燃气轮机建模仿真基础上,以燃气轮机排气温度为例,对时间序列模型在燃气轮机气路性能退化预测中的应用进行了有效性评估。研究结果表明,使用该方法可将燃气轮机气路性能参数随压气机积垢程度的变化转化为时间序列趋势预测问题,可以有效地实现燃气轮机气路性能退化趋势预测,进而实现压气机的视情清洗维护。该方法为燃气轮机在线健康状态监测与诊断提供了一种思路,具有一定的工程应用价值。

For the problems of compressor performance degradation and regular cleaning maintenance caused by compressor fouling in the actual operation process of gas turbine,the paper develops a prediction approach based on time series model for gas path performance degradation of gas turbine. On the basis of gas turbine thermal modeling and simulation,taking exhaust gas temperature of gas turbine as an example,the paper conducts the validity evaluation on the application of time series model in the prediction on gas path performance degradation of gas turbine. The study results indicate that by using the approach,the prediction on the changes of the gas path performance parameters of gas turbine can be transformed into the time series prediction problem. In this way,the prediction on gas path performance degradation of gas turbine can be realized effectively,and then cleaning maintenance as appropriate for compressor can also be implemented. The approach provides a new way to online condition monitoring and fault diagnosis for gas turbine,so it has a certain engineering application value.