研究目的
Investigating the improvement of photovoltaic power generation forecasting using Multiple Linear Regression method with a real-time correction term.
研究成果
The MLRM+RCT model significantly improves the prediction accuracy of PV power generation, benefiting the stable operation of PV power plants and the power grid. Future work includes optimizing the selection of characteristic variables and using adaptive methods to identify RCT parameters.
研究不足
The study relies on historical data and may not account for all real-time variations in weather conditions. The selection of characteristic variables and interaction items could be optimized further.
1：Experimental Design and Method Selection:
The study designs a day-ahead, hourly PV power generation prediction model using MLRM, considering qualitative and quantitative variables and their interactions. A real-time correction term (RCT) is added to MLRM to improve forecasting accuracy.
2：Sample Selection and Data Sources:
The power generation data from three photovoltaic power stations in Australia, released by IEEE Energy Forecasting Group in 2014, is used. Meteorological data is sourced from the European Centre for Medium-Range Weather Forecasts.
3：List of Experimental Equipment and Materials:
MATLAB 2018a is used for estimating regression coefficients of MLRM.
4：Experimental Procedures and Operational Workflow:
The model is trained with data from one PV power station and tested with another set. The RCT is applied to correct forecasting results based on real-time measured power generation data.
5：Data Analysis Methods:
Performance is evaluated using mean error (ME), mean absolute error (MAE), mean squared error (MSE), and mean absolute percentage error (MAPE).
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
