Probe into micro-grid short-term load forecasting based on method of chaotic time series

The paper employs chaos theory in electric power system forecasting. The chaos prediction model is established via phase space reconstruction, and a new improved local prediction method is proposed and applied in micro-grid short-term load forecasting. Delay time and embedding dimension are calculated to reconstruct the phase space. The solved maximal Lyapunov exponent reveals the chaotic characteristics of the load time series, and the characteristic difference between the load time series and random series is analyzed through waveform comparison. Historical load data length and sampling interval of daily load are taken into consideration. Daily sampling intervals were taken as 15min, 30min, 45min and 60min, respectively. Finally, the model is applied to forecast actual load data of a certain community in Shaanxi province. Reasonable value range of the data length and sampling interval are given and the results can be taken as one reference for the optimal operation of the actual power grid.