초록
풍력발전설비는 접근성의 문제로 주기적인 구조건전성 검사를 수행하기 어렵고, 기상상태를 포함한 주위 환경변화 때문에 예기치 못한 고장발생 가능성이 높아 이에 대한 보완책으로 상태감시시스템(Condition Monitoring System, CMS)을 운영하고 있다. 본 연구에서는 CMS의 이상감시 성능 향상을 위하여 풍력발전기 주요 기계시스템에서 장기간 측정된 진동신호 분포를 통계적으로 분석하고, 운전 조건에 따른 진동 변화 경향을 파악한다. 이를 위하여, 풍력발전기 동력전달 및 전력생성부의 진동, 풍속, 주축회전수 등을 약 2년동안 측정한 데이터를 기반으로 운전 환경 및 조건에 따른 각 신호의 경향분석을 수행하고, 기계시스템 구조에 따른 신호별 상호연관성을 분석하였다. 결과적으로 풍력발전기 기계시스템별 진동은 주축회전수, 발전여부에 영향을 받고, 특정 주축회전수에서는 베이불(Weibull) 분포에 해당하는 진동분포가 나타남을 확인하였다. 이런 결과는 풍력발전기 CMS 시스템에서 기계적 이상발생 여부를 조기에 판단하는 기준을 제시할 수 있다.
Condition Monitoring System (CMS) has been used to detect unexpected faults of wind turbine caused by the abrupt change of circumstances or the aging of its mechanical part. In fact, it is a very hard work to do regular inspection for its maintenance because wind turbine is located on the mountaintop or sea. The purpose of this study is to find out distribution patterns of vibration signals measured from the main mechanical parts of wind turbine according to its operation condition. To this end, acceleration signals of main bearing, gearbox, generator, wind speed, rotational speed, etc were measured through the long period more than 2 years and trend analyses on each signal were conducted as a function of the rotational speed. In addition, correlation analysis among the signals was done to grasp the relation between mechanical parts. As a result, the vibrations were dependent on the rotational speed of main shaft and whether power was generated or not, and their distributions at a specific rotational speed could be approximated to Weibull distribution. It was also investigated that the vibration at main bearing was correlated with vibration at gearbox each other, whereas vibration at generator should be dealt with individually because of generating mechanism. These results can be used for improving performance of CMS that early detects the mechanical abnormality of wind turbine.
