Analysis Of Variable Condition Property Of Governing Stage Based On Actual Operating Data

With the rapid development of the national economy, the power consumption structure of China has changes significantly. The peak valley difference of national power grid load curve also increases gradually, peak regulation of power network causes that the unit is often at off-design conditions even depth off-design conditions. When the Steam turbine is under varying conditions, the steam flow rate, the steam pressure of the nozzle and the moving blades, humidity, steam temperature, efficiency, enthalpy drop of the stage will deviate from the original design value, and the variation of parameters provides an important scientific basis for failure prediction, axial thrust calculation, and comprehensive understanding of the unit’s efficiency and power changes. Off-design calculation of stage provides the basis of the off-design calculation of the steam turbine, while the difficulty of the stage calculation is the characteristic analysis of the governing stage and the last stage.Based on actual operating data of turbine, through series of processing method such as steady-state judgment and abnormal data processing, after the procedural of the program, Batch processing of large amounts of data will be realized. Then select the data which satisfy intelligent module to compute data requirements of SIS platform, optimize the environment of the application of intelligence platform. After data preprocessing, by analyzing the sequential method applied in the calculation of the characteristic analysis of the governing stage, looking for main factors that influences the characteristic of governing stage, analyzing the advantages and disadvantages of the sequence method in the characteristic analysis of the governing stage, eventually puts forward improving methods based on certain theories. After finding out the strong correlation parameters of unit power, main steam pressure, temperature, temperature after regulating stage, opening of each control valve, and pressure after regulating stage, establish a statistical model based on support vector machine(SVM), validate by testing data, analyze calculated error, model stability, and Whether the model has good applicability. According to the result feedback, readjust the parameters of the model set, eventually meet accuracy before the system integration. Based on actual operating data, software developed by Delphi to analyze the characteristics of the governing stage off-design condition is of great significance for guiding efficient operation.