발전량 데이터 분석을 통한 신재생에너지 주택지원사업지의 태양광발전 시스템 이상 감지

Abstract

In this study, we proposed a new method to detect the abnormalities among photovoltaic (PV) systems installed in renewable energy housing support project sites by analyzing power generation data. The site at the north side of Gakbuk-myeon, Cheongdo-gun, Gyeongsangbuk-do, Korea was selected as a study area where 63 PV systems have been installed and operated. By considering the factors related to system design and surrounding environment, the 63 PV systems were clustered into 6 groups using the K-means clustering method which is an unsupervised machine learning algorithm. The power production data from the PV systems for each group was analyzed and set as an abnormal value if it deviates from the range of ±2.58 times the standard deviation from the mean (assuming a normal distribution and 99% confidence level confidence interval). As a result, several abnormalities among PV systems were detected during November 2020. The cause of abnormalities was confirmed through a site investigation. It is expected that the proposed method can be effectively used for fast problem diagnosis of PV systems in renewable energy housing support project sites.