Key points are not available for this paper at this time.
In recent years, with the rapid development of technology, the level of informatization of power systems based on cloud computing has also been continuously improving. More and more data has been accumulated in various power grid systems, forming power big data. In previous smart grids, grid load forecasting was a crucial step, and the design of its prediction algorithm was also a crucial issue. The accuracy of power grid load forecasting results not only affects the normal operation of the entire smart grid, but also affects the macroeconomic regulation of the power grid. Therefore, analyzing and mining big data on electricity plays an important role in the construction of smart grids and the efficient, safe, and stable operation of the power grid. Abnormal detection and prediction of power load is an important research topic in power big data mining. This article studies the method of detecting and predicting abnormal power loads and implements it in a cloud computing environment. After optimization, the average error of neural network prediction is basically controlled within 3%, which is sufficient to meet the requirements of power system prediction and create higher benefits for society. It has certain practical value.
Zhang et al. (Fri,) studied this question.