Forecasting of PV Power Output Based on SupportVector Regression and Fuzzy Inference Approach

• Main Authors: Yi-ShiangPai, 白亦翔
• Other Authors: Hong-Tzer Yang
• Format: Others
• Language: en_US
• Published: 2013
• Online Access: http://ndltd.ncl.edu.tw/handle/39723167198812380633

碩士 === 國立成功大學 === 電機工程學系碩博士班 === 101 === This thesis uses support vector regression (SVR) and fuzzy inference method for one-day ahead forecasting of photovoltaic (PV) power output. SVR employed in this thesis has been successfully applied to data classification and regression analysis. It uses the best hyperplane to extract features from linear or nonlinear data. In the training stage, the SVR is trained by using the collected input data for temperature, probability of precipitation, solar irradiance of defined similar hours, which are classified via fuzzy inference method. In the forecasting stage, the fuzzy inference method is used to select an adequate trained model according to the weather information collected from Taiwan Central Weather Bureau (TCWB). The proposed approach is applied to a practical PV power generation system. This thesis uses one-year weather information collected from TCWB to test the PV power forecasting. The comparison with the actual data is used to verify the accuracy. Numerical results show that the proposed approach achieves better prediction accuracy than the simple SVM method and traditional ANN method.