An Detection Method for Steam Turbine Flow Path Fouling Based-on Support Vector Regression

The fouling on flow path has a great influence on the safety and economic of steam turbine. In the paper, the limitations of main steam flow measurement model based-on Fulugel formula are analyzed firstly, then countering the shortage that the main steam flow measurement model based-on Fulugel formula is inaccurate, a model is structured based-on Support Vector Regression, and comparing the regression results with measurements to determine whether the flow path section of steam turbine is fouling. The calculation result indicates that the model based on Support Vector Regression has not been affected by the area change of flow path section. It provides a new calculation method of main steam flow and analysis method of flow path section.