Prediction of electric power consumption by nuclear power plants for covering plant needs using data mining methods

Nuclear Energy and Technology

关键词: Nuclear power plants;    Energy consumption;    Forecasting methods;    Exponential smoothing;    Support vector machines;   

DOI  :  10.1016/j.nucet.2016.02.010

来源: DOAJ

【 摘 要 】

In the process of its operation nuclear power plant consumes significant amounts of electric power – the so-called power consumption for coverage of plant needs. According to the practice existing in Russia, nuclear power plant must order in advance from the power supply grid operator the required quantity of electric power, while deviations of factual electric power consumption from the forecasted value one way or another entail payment of certain financial penalties. Because of this reason accuracy of forecasting the electric power consumption for the nearest coming period attains special importance. Application of different methods of data analysis for forecasting electric power consumption by NPPs using factual data is addressed and comparison of these methods between each other, as well as with those currently in application by NPPs is made. As the result, a method is suggested for forecasting NPP power consumption for coverage of plant needs with significantly higher accuracy of forecasting.

【 授权许可】

Unknown