Kai Wu / China Nuclear Power Technology Research institute
Hetao Sun / Harbin Engineering University
Under normal shutdown or accident conditions, the control rod should be dropped within a predetermined time to achieve safe shutdown. Because of the special working principle of the control rod hydraulic drive system, the accurate drop curve is affected by many pressure and flow parameters. However, because of the uncertainty of the measuring device, it is difficult to get the accurate drop curve in the experiment. By establishing the BP neural network model and preprocessing the input data of the network with the statistical method, the control rod fast falling curve is predicted.The results show that the method not only overcomes the uncertainty of the measuring device, but also can quickly feed back the drop curve and adjust the pressure and flow of the driving system accurately and timely.