Wind Turbine Gearbox Condition Monitoring Vibration Analysis Benchmarking Datasets

Wind turbine condition monitoring (CM) can potentially help the wind industry reduce turbine downtime and operation and maintenance (O&M) cost. NREL CM research has investigated various condition-monitoring techniques such as acoustic emission (AE specifically stress wave), vibration, electrical signature, lubricant and debris monitoring based on the Gearbox Reliability Collaborative dynamometer and field tests, and other test turbines and resources accessible by NREL. During the past several years, NREL CM research has shown that there are very few validation and verification efforts on commercial wind turbine CM systems. One of the reasons might be limited benchmarking datasets accessible by stakeholders. To fill this gap, NREL executed a data collection effort. The targeted users of these datasets include those investigating vibration-based wind turbine CM research, evaluating commercially available vibration-based CM systems, or testing prototyped vibration-based CM systems. NREL collected data from a healthy and a damaged gearbox of the same design tested by the GRC. Vibration data were collected by accelerometers along with high-speed shaft RPM signals during the dynamometer testing. The healthy gearbox was only tested in the dynamometer. The damaged gearbox was first tested in the dynamometer and later sent to a wind farm close to NREL for field testing. In the field test, it experienced two loss-of-oil events that damaged its internal bearings and gear elements. The gearbox was brought back to NREL and it was retested in the dynamometer with CM systems deployed under controlled loading conditions that would not cause catastrophic failure of the gearbox. The objective of releasing these datasets to the public along with information about the real damage that occurred to the damaged gearbox is to provide the wind industry with some benchmarking datasets. These datasets will benefit research, development, validation, verification, and advancement of vibration-based wind condition-monitoring techniques. By accessing this data you acknowledge the terms outlined in the "License Information" document. Please contract Shawn Sheng (NREL) if you have any questions on the data or would like to collaborate on publications based on the datasets.

风力发电机组状态监测（Condition Monitoring，CM）可为风电行业降低机组停机时间与运维（Operation and Maintenance，O&M）成本提供助力。美国国家可再生能源实验室（National Renewable Energy Laboratory, NREL）的状态监测研究团队，依托齿轮箱可靠性协作项目（Gearbox Reliability Collaborative, GRC）测功机试验、现场测试，以及NREL可获取的其他测试涡轮机与相关资源，对多种状态监测技术开展了研究，涵盖声发射（Acoustic Emission, AE，特指应力波）、振动、电信号、润滑油及磨粒监测。过往数年间，NREL的状态监测研究显示，针对商用风力发电机组状态监测系统的验证与确认工作仍极为稀缺，究其原因之一，或许是利益相关方可获取的基准数据集有限。为填补这一研究空白，NREL启动了数据采集工作。本数据集的目标用户包括开展基于振动的风力发电机组状态监测研究、评估商用振动式状态监测系统，或测试原型振动式状态监测系统的科研与工程人员。NREL从GRC测试的同型号健康齿轮箱与故障齿轮箱中采集了相关数据。测功机测试期间，通过加速度传感器与高速轴转速信号同步采集振动数据。其中健康齿轮箱仅在测功机环境下完成测试，而故障齿轮箱先在测功机中完成测试，随后被运送至NREL附近的风电场开展现场测试。在现场测试阶段，该齿轮箱先后经历两次缺油事件，致使其内部轴承与齿轮部件受损。之后该故障齿轮箱被运回NREL，在受控载荷条件下再次于测功机中开展复测，该载荷设置不会引发齿轮箱发生灾难性失效。公开本数据集及故障齿轮箱实际受损情况的相关信息，旨在为风电行业提供可供参考的基准数据集。此类数据集将为基于振动的风力发电状态监测技术的研究、开发、验证、确认及技术迭代提供有力支撑。使用本数据集即视为您已同意遵守“许可信息”文档中列明的各项条款。若您对本数据集存在任何疑问，或希望基于本数据集合作发表学术论文，请联系NREL的Shawn Sheng。

搜集汇总

背景与挑战

背景概述

该数据集包含健康和损坏风力涡轮机齿轮箱的振动监测数据，旨在支持振动状态监测技术的研发和验证，为行业提供基准测试资源。