A novel interpolation method for high temporal-resolution electricity usages based on volatility quantification and pattern extraction

As the penetration rate of renewable energy in the power grid increases, it is essential to reduce energy consumption and carbon emissions associated with buildings. Interactions between a building’s microgrid and power grids have attracted significant research attention. However, most building energy data have been collected at 1-hour intervals, which mismatches the 15-min intervals used in current power grid control and demand responses. Variations in building electricity data at different temporal time-steps were observed in electricity load distribution, daily peak usage, and volatility values, and existing high-resolution interpolation methods failed to quantify these features. In this study, we developed a novel high temporal-resolution interpolation method for hourly building electricity usage data, using volatility quantification and pattern extraction. The proposed methodology includes two folds: analyzing daily fluctuation amplitudes and analyzing hourly fluctuation patterns. Validation results of the proposed method on high-resolution electricity data of 19 commercial complexes in North and Northeast China was conducted, and compared to a baseline zero-order interpolation model. The proposed method significantly decreased the median KS statistic from 0.0155 to 0.0048 for electricity load distribution, from 0.0712 to 0.0356 for daily electricity peak usage distribution, and from 0.6932 to 0.1726 for daily electricity VF value distribution. These results indicate that the proposed method accurately captures the critical features of high-resolution building electricity usage profiles and improved the quantification of electricity usage volatility and randomness. Additionally, the proposed method enhanced the precision of the analysis of operational energy costs and battery cycles, supporting the design and operation of the building microgrid systems.