Study of the modeling of hydroturbine generating set

To design or operate the hydroturbine governing system of hydroelectric power plant the dynamic model of hydroturbine generating set is needed, but it is hard to obtain it theoretically. Using system identification and parameter estimation is effective. Based on the field test of real machine set and using recursive least-square estimation (RLES) algorithm, the authors approach the selection of exciting signal parameters and sampling period; the relationship between matrix P and identified parameters convergence; the interaction of forgetting factor /spl lambda/ and the positive definite matrix P etc. The paper emphasises practical problems with real machines. The modeling of dynamic model between the opening of guide vane (input) and the rotating speed of hydroturbine generating set (output) is a very important topic in the design or operation of hydraulic turbine governing system, because it decides the regulation parameters of the governor, specially at the nonload state.