Abstract: For the purpose of conducting real-time analysis and online monitoring on the valve discharge characteristics of the steam turbine, one dynamic data mining model is built up for the valve discharge characteristics calculation. A 350 MW turbine set is used to validate the dynamic data mining model. The results show that the dynamic data mining model can extract and update calculation samples. Moreover, the mathematical model established by moving least square method can simulate the mapping relationship between the input and output parameters well. The average relative error is less than 1.5% between the mathematical model and the results of the steam turbine performance testing, which verifies that this dynamic data mining model is an effective approach for valve discharge characteristics online calculation.

Key words: thermal power plant, steam turbine, valve discharge characteristics, dynamic data mining, moving least square