研究目的
To forecast photovoltaic (PV) power generation using Kalman filter and Auto Regressive Integrated Moving Average (ARIMA) for real-time forecast with high resolution time step, and to propose an estimator model for reliable forecast when real-time measurement data is unavailable.
研究成果
The ARIMA-Kalman filter approach with an estimator model provides a reliable method for forecasting PV power generation, especially when real-time measurement data is unavailable. The estimator model, based on neighbor PV rooftops' data, supports the ARIMA-Kalman approach effectively.
研究不足
The accuracy of the forecast depends on the availability and quality of historical data from neighbor PV rooftops. The model may require adjustment for different geographical locations or weather conditions.
1：Experimental Design and Method Selection:
The study uses ARIMA model to estimate transition matrix for running Kalman filter. The estimator model uses historical data of power generation from neighbor PV rooftops and distance between PV rooftops.
2：Sample Selection and Data Sources:
Historical data from three PV rooftops located near each other are used. Data are inspected and fixed for missing data by linear interpolation.
3：List of Experimental Equipment and Materials:
MATLAB software is used to run Kalman filter algorithm.
4：Experimental Procedures and Operational Workflow:
The process involves determining parameters for ARIMA model, estimating coefficients, and using Kalman filter for forecast. The estimator model is used when real-time data is unavailable.
5：Data Analysis Methods:
The performance is measured by Root Mean Square Error (RMSE) and Skill Score (SS).
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
