研究目的
To present a new spatial–temporal forecasting method for solar power forecasting in smart grids, aiming to improve the accuracy of 6-h-ahead forecasts at the residential solar photovoltaic and medium-voltage (MV)/LV substation levels.
研究成果
The spatial–temporal forecasting method based on the VAR framework significantly improves the accuracy of solar power forecasts for smart grids, with an average improvement of 8% to 10% over the autoregressive model. The method leverages data from distributed PV generation to enhance forecast accuracy, particularly for the first three lead times.
研究不足
The proposed method is suitable for very short-term horizons (up to 6 h ahead) and may not perform as well for longer time horizons without the inclusion of numerical weather predictions or satellite information.
1：Experimental Design and Method Selection:
The study employs a vector autoregression (VAR) framework for spatial–temporal forecasting of solar power, combining observations from smart meters and distribution transformer controllers.
2：Sample Selection and Data Sources:
Data from 44 microgeneration units and 10 MV/LV substations in the smart grid pilot of évora, Portugal, were used.
3：List of Experimental Equipment and Materials:
Smart meters and distribution transformer controllers were utilized for data collection.
4：Experimental Procedures and Operational Workflow:
The VAR model was applied to forecast solar power for each residential PV and secondary substation, with a benchmark comparison made against an autoregressive forecasting model.
5：Data Analysis Methods:
The forecasting accuracy was evaluated using the root-mean-square error (RMSE), normalized with the solar peak power.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
