Statistical Characteristics and Non-Gaussianity of Grid Frequency Based on Measured Data

The grid frequency is a crucial factor in ensuring the stable operation of the power system and, therefore, must be strictly maintained close to the nominal frequency. With the integration of renewable energy generation into the grid, the impact of more frequent power variations on power quality intensifies. In this context, studying the statistical properties of frequency holds significant guiding importance. In this paper, we gathered real-time grid operation data from SML devices deployed nationwide. We analyzed the mean, variance, and other operational indices for the 24-hour frequency data of each location on the same day. Furthermore, we calculated the autocorrelation of the frequency data and identified the implied trading interval of the electricity market. We also investigated the non-Gaussianity of the frequency probability distribution in China's power grid. Our findings indicate that primary frequency control, secondary frequency control, and load power all influence the formation of the bimodal distribution, a conclusion substantiated through simulation experiments. The results of this study can provide guidance for controlling and researching applications based on grid frequency.