Study On Control Method Of Sluice Gate In Middle And Small Hydropower Station Based On Neural Network

Gate automatic control device is the core part of reservoir and hydropower station. There are the following defects under manual control valve.In the face of the complex water lines, the opening of the gate can’t be precise control; Lock water flow can’t be precise calculating;It has the problem of low efficiency and poor accuracy. It is necessary to automatically control the gate of hydropower station using the modern control method combined with computer network communication technology.It cannot only improve the control accuracy, prevent danger to avoid safety accident, also can improve the efficiency of work.In this paper, we analysis and research running status of the hydropower station gate for YaoZhuang in Hebei FengRun.This hydropower station is a medium-sized hydropower station.It’s installed capacity is 2500 Kw.Water flow can reach 35 cubic meters per second.The output voltage is minus 15 kV. The hydropower station is put into use in 1995.Currently it can only control the gate using artificial operation way.This way of control precision and accuracy is lower, for more rain in the summer and the operation of some special cases.Also it increases unnecessary consumption, and it is bad for the present stage production use.To overcome defects above, it requires automatic control the gate of hydropower stations.In this paper,it adopts the traditional PID control for the master, using BP neural network to adjusti PID parameters in real-time to solve the main algorithms in the drain valve’s automatic control system.and research methods of RBF neural network control and BP neural network PID control parameters for further. Finally, we carry out simulation experiment by MATLAB.Research shows controlling PID parameters by RBF neural network, the response speed and convergence effect is superior tocontrol PID parameters by the BP neural network, but by contrast structure is too complex.Due to the small and medium sized hydropower stations to adjust accuracy requirement is not very high, and not suit to use overly complex equipment. Therefore, it is the most suitable control system of automatic algorithm for the small and medium-sized hydropower station water gate choosing the BP neural network.