基于指数加权移动平均的燃气轮机参数特征检测方法OA

The gas turbine health monitoring system is used for monitoring,isolating and diagnosing faults of main components in gas turbine.Step and ramp change detection are the foundations of health monitoring,and the results can reflect abnormal operation of gas turbine.By analyzing the shortcomings of conventional detection technologies of step and ramp features,a detection algorithm for gas turbine parameter feature based on statistical methods was proposed to automatically detect potential step and ramp changes of the monitored parameters.Actual test data of gas turbine validated that the algorithm could effectively detect abnormal step and ramp changes in parameters,thus,it could assist maintenance personnel to monitor the health status of gas turbine.The research findings can provide reference for optimization of health monitoring technology for gas turbine.