A data cleaning method based on spatiotemporal inverse distance weight interpolation for meteorological data in power system

Abstract

The scale of modern power grids is huge and the structure is complex, and most of the power equipment is directly exposed to the external environment. In recent years, the problems of global warming and environmental damage have become increasingly prominent. The impact of disasters on the safe and stable operation of power grids has become increasingly prominent. In order to deal with the impact of meteorological disasters on the power grid and reduce the losses caused by meteorological disasters to the power system, it is necessary to build a micrometeorological monitoring station for the power grid, and use the meteorological monitoring data for disaster early warning analysis. However, existing meteorological monitoring stations have a series of data quality problems due to the particularity of their own spatial locations, and there are a lot of dirty data in the original data, which seriously affects the accuracy of subsequent data analysis. To this end, this paper proposes a power meteorological data cleaning method based on spatiotemporal inverse distance weight interpolation. Experiments show that this method can remove most of the dirty data in the original power meteorological data, thereby effectively improving the prediction accuracy of power grid meteorological disasters. Safe and stable operation is of great significance