ANN Based Modeling and Optimization of Large Pumped Storage Station

Modeling of regulating system of large pumped storage power station with long pipeline and parameters optimization in transient processes are studied in this paper. According to the actual parameters of a certain pumped storage power station, MATLAB/Simulink toolbox is utilized for modeling and simulation of its subsystems. Friction resistance coefficient and water elasticity are taken into consideration in modeling of pressure diversion syste。As to simulate hydraulic vibration characters, BP neural network and RBF neural network are adopted in modeling of pump turbine. Based on the established regulating system simulation model, improved orthogonal experiment method is applied in parameters optimization of no-load frequency disturbance, load disturbance and load shedding transient processes. According to the results, the proposed model reflect the actual characteristics of pumped storage units, and improved orthogonal experiment method is effective in figuring out the optimal parameters group within the given range. This paper provides guidance for modeling of regulating system of large pumped storage units, and set references and theoretical basis for actual optimal control of transient processes in pumped storage units.