| It is required to monitor and forecast the running status and the long-term trend of the machinery in fieldwork for the machines' working safely and steadily in long period and full-time. Implement to manage and maintain the equipments in advance according to the equipment conditions, which prevents accidents before they occur, reduces the breakdown suddenly and the maintenance cost, brings the economic efficiency for the enterprise and the society. The tendency forecast of equipment status is one of the important parts for modern equipment dynamic management, especially for very complex systems such as The Nuclear Power Plant (NPP) which has serious consequence and big large-scale service cost, moreover the badly negative society affects once faults happen.This paper formulates the functions of fault diagnosis system, and refocuses in the tendency forecast of equipment status according to the characteristics of faults diagnosis of NPP. Taking the Daya Bay Nuclear Power Station as an example, this paper also sorts the different running status of equipments and collects abnormity signals and character signals of the main faults of every systems of the primary loop. All of these are necessary to develop the tendency forecast of equipment status system.The process is very complex and the forecast result may not be satisfied in traditional forecasting methods. Especially, traditional methods need lots of data, which is almost impossible for the NPP. Gray Forecast Models can be used to establish models to forecast with less data. Owing to the randomness and fluctuation of the data from NPP, the data fitting is difficult and precision is low. This paper improves the model by considering these factors, and brings forward Gray-Markov forecast models. And we may obtain the more precise forecast result by using optimization and feedback adjustment and others in the model. In the end of this paper, the forecast model is applied in NNP, which gets good result.
|
PDF Full Text Request