Abstract:The intermittent distributed generation,along with electric vehicles and electricity market prices bring out frequent load fluctuation and impose great operational uncertainties on the distribution networks. Long time series analysis of load provides a more comprehensive manner for characterizing consumption patterns of electricity consumers, which greatly improves robustness of analysis results. However, little research pays attention to the long load time series and its fluctuation complexity. Inspired by the application of complex network in nonlinear time series analysis, we transform the long load time series into multiscale networks and explore fluctuations complexity of the original long load time series by relevant network topological parameters from the view of statistical analysis. To explore predictability of a single consumer and dispatchability among consumers, the experiments on several successive time intervals of load fluctuations analysis for a single consumer and load fluctuations analysis for multiple consumers are conducted in this paper. According to the numerical experiments, the proposed method can implement the quantitative analysis of load fluctuations complexity for long load time series.

Revealing the Complexity of Load Fluctuations for Electricity Consumers Using MLP-HVG