REAL-TIME CONDITION MONITORING OF OFFSHORE WIND TURBINES
REAL-TIME CONDITION MONITORING OF OFFSHORE WIND TURBINES
复制标题
海上风力发电机的实时状态监测
DOI：
--
复制
发表时间：
2006
期刊：
影响因子：
--
通讯作者：
J. Xiang
中科院分区：
文献类型：
--
作者：
S. Watson;J. Xiang
翻译后摘要：状态监测已被广泛用于在电力行业中，以监测传统发电机的性能。这种发电机经常在恒定负载下长时间运行，并且由于潜在的故障模式而引起的点变化使用标准信号处理技术是相对简单的。风力涡轮机的状态监测的应用相对较新，并且依赖于相对粗糙的警报水平的设置，高于该警报水平，可以在几乎没有关于故障模式或具有足够警告的可能故障时间的信息的情况下发起停机。许多现代风力涡轮机的变速运行对“智能”状态监测系统的应用既是挑战也是机遇。一方面，可变载荷意味着难以在频谱内发现潜在的失效模式。另一方面，可变负载可以激发风力涡轮机内的模式范围，并且潜在地阐明关于涡轮机的健康的重要信息。在近海环境中，对具有关于特定故障模式的更精确信息和对平均故障时间的准确预测的有效状态监测的需求变得越来越迫切。在本文中，作者提出了一种新的方法，利用快速傅立叶变换（FFT）和小波的优点，并说明其应用于检测的发电机轴承故障在1.5 MW变桨距变速风力涡轮机。
Abstract: Condition monitoring has been used extensively in the power industry to monitor the performance of conventional power generators. Such generators are frequently run for long periods at constant loading and spotting changes due to potential failure modes is relatively straightforward using standard signal processing techniques. The application of condition monitoring to wind turbines is relatively new and relies on the setting of relatively crude alarm levels above which a shut-down may be instigated with little information about the failure mode or possible time to failure with sufficient warning. The variable speed operation of many modern wind turbines represents both a challenge and an opportunity to the application of an ‘intelligent’ condition monitoring system. On the one hand a variable loading means that it can be difficult to spot a potential failure mode within a frequency spectrum. On the other hand, a variable load can excite a range of modes within a wind turbine and potentially elucidate significant information about the health of the turbine. The need for effective condition monitoring with more precise information about a particular failure mode and accurate prediction of mean time to failure becomes ever more acute within the offshore environment. In this paper, the authors present a new method that utilises the advantages of both fast fourier transforms (FFTs) and wavelets and illustrate their application to the detection of a generator bearing fault in a 1.5MW pitchregulated variable speed wind turbine.