Prediction of renewable energy hosting capacity using multiple linear regression in KEPCO system
Kyungsang Lee, Seunghyuk Im, Byongjun Lee
Abstract
In order to prepare measures for the stable operation of the power system to expand renewable energy, the renewable energy hosting capacity (HC) in the system shall be identified in advance. This paper proposes a methodology for predicting monthly HC based on factors affecting HC. It was found out that these factors are: Total generation, ratio of nuclear, coal, liquefied natural gas (LNG), and other power generations. A prediction model was developed using multiple linear regression by integrating and separating data of elements from weekend data. A comparison of the determination coefficients showed that the models incorporating weekend data exhibited the best accuracy. In conclusion, the proposed model has the characteristics of predicting various HCs simply and quickly with five factors.