인문학
사회과학
자연과학
공학
의약학
농수해양학
예술체육학
복합학
DBpia AI
AI 에이전트
AI 아이디어
AI 리더
AI 뷰어
Citeasy
크롬 서지관리 다운로드
지원사업
학술연구/단체지원/교육 등 연구자 활동을 지속하도록 DBpia가 지원하고 있어요.
커뮤니티
연구자들이 자신의 연구와 전문성을 널리 알리고, 새로운 협력의 기회를 만들 수 있는 네트워킹 공간이에요.
피인용 0
초록·키워드
The performance evaluation and operation and maintenance (O&M) of photovoltaic (PV) power plants are primarily based on key performance indicators (KPIs) derived form monitoring data, including Performance Ratio (PR) and Capacity Factor (CF). In this study, we evaluated the feasibility of applying unsupervised learning-based toward anomaly detection using daily KPIs generated from conventional 10-min interval monitoring data. Daily CF and temperature-corrected PR were calculated from approximately 4.5 years of operational data collected from an 88.2 kWp PV power plant and used as input features for Isolation Forest (i-Forest), k-means, and Gaussian Mixture Model (GMM). We found that CF is suitable for long-term performance assessment but is limited in its capacity to detect anomalies, whereas PR is highly sensitive to performance degradation but is also susceptible to irradiance sensor measurement errors. Therefore, CF and PR facilitates the characterization of both degradation-related anomalies and measurement-related abnormalities. Among the evaluated models, GMM exhibited the best clustering performance by clearly identifying normal operation, performance degradation, low-irradiance operation, and irradiance sensor anomalies. These results demonstrate the feasibility of using daily KPIs derived monitoring data for the unsupervised identification of abnormal operating patterns. This study highlights the feasibility of developing practical and cost-effective AI-based O&M technologies for PV power plants using existing monitoring infrastructure.
본문·목차
인공지능 문자 인식 모델을 통해 추출된 텍스트로, 일부 오타나 오류가 포함될 수 있으나 지속적으로 개선 중입니다.
오류를 발견하셨다면 해당 부분을 드래그한 후 ' 를 통해 신고해주세요.
오류를 발견하셨다면 해당 부분을 드래그한 후 ' 를 통해 신고해주세요.
최근 본 자료 전체보기
