Development and demonstration of methods and tools for large scale wind turbine pitch bearing condition assessment (DemoBearing)

英国是世界上海上风力发电装机容量最大的国家，并将在未来几十年继续以优势速度部署，以实现2050年的碳排放目标。海上风力涡轮机的尺寸越来越大，对其所有部件的操作和维护提出了重大挑战。特别是风力发电机桨距轴承，作为风力发电机叶片与轮毂之间的安全关键接口，用于旋转叶片实现发电优化和紧急停机，是典型的大、慢、部分旋转轴承，但却是大型海上风力发电机的薄弱部件和瓶颈（新兴的重大挑战）。此外，到2030年，英国将有大量陆上涡轮机接近其设计寿命。螺距轴承对老化涡轮机退役或延长寿命的决策构成重大风险（即将面临的挑战）。现场节距轴承状态评估是整个风电行业面临的一个重大挑战，因为目前还没有行业标准，而且现有的原位方法（如内窥镜检查和油脂分析）只能部分评估节距轴承状况。因此，开发有效的现场状态评估方法和工具至关重要，以降低高昂的维护成本、计划外停机时间和灾难性故障风险，提高陆上和海上风力发电的可靠性和能源效率，并为老化的陆上风力涡轮机延长寿命提供可靠的决策。这项雄心勃勃的研究是首次在国际前沿开发智能音高轴承状态评估方法和使用振动和声发射测量的原位工具。特别是，该研究通过解决与大型慢速轴承的弱、噪声和非平稳数据分析相关的根本技术挑战，解决了风电行业的全球重大挑战。这将通过开发具有稀疏信号分离、数据融合和机器学习方法的新算法来实现，然后在实验室和现实世界的操作环境中进行重要的演示活动。PI开发了第一个工业规模的风力涡轮机螺距轴承平台，其中包括三个自然损坏的轴承，在实际风电场中使用寿命超过15年，以及先进的数据收集仪器。新搭建的平台为拟开展的研究奠定了坚实的基础，为开展示范和影响活动创造了理想的平台。该项目还获得了独特的机会，在两个工业项目合作伙伴提供的最有力支持下，在实际运营的风电场中进行现场数据收集和演示。从三个自然损坏的轴承收集的数据将在开源许可下公开提供，使其他研究人员能够对大型慢速轴承进行状态评估。项目期间开发的知识产权将受到保护。如果不与知识产权冲突，开发的算法将向公众开放。该项目的成功成果将在原位节距轴承状态评估方法和工具方面开辟新路，有助于节距轴承的工业标准，并在数十年的轴承使用寿命中受益于使用大型慢速轴承的广泛行业，如海上石油，天然气，采矿和钢铁制造。这种针对弱、噪声和非平稳数据分析的新方法可用于广泛的数据驱动应用。因此，该项目在未来几十年将产生重大、广泛和长期的影响。

The UK is No. 1 in the world for installed offshore wind power and continues the deployment in a predominant speed in the next few decades to meet 2050 carbon emissions targets. The increasing sizes of offshore wind turbines pose significant challenges in the operation and maintenance of all its components. In particular, wind turbine pitch bearing, as the safety-critical interface between the turbine blade and the hub to rotate the blade for power generation optimisation and emergency stop, is typified as the large, slow, partially rotated bearing but it is the weak part and bottleneck for large offshore turbines (Emerging grand challenge). In addition, the UK will have a large number of onshore turbines approaching the end of their design life by 2030. The pitch bearing poses a significant risk for the decision making in ageing turbine decommissioning or life extension (Upcoming challenge). In-situ pitch bearings condition assessment is a major and open challenge for the whole wind industry as there are no industrial standards available yet and few existing in-situ methods, such as endoscopy and grease analysis, can only partially assess the pitch bearing conditions. Therefore, it is essential to develop effective in-situ condition assessment methods and tools in order to reduce high maintenance cost, unplanned downtime and risk of catastrophic failure, improve reliability and energy efficiency of onshore and offshore wind power generation and enable reliable decision making in ageing onshore wind turbine life extension.The ambitious research is, for the first time and at the international forefront, to develop intelligent pitch bearing condition assessment methods and in-situ tools using vibration and acoustic emission measurements. In particular, the research tackles the global grand challenges in wind industry by addressing the fundamentally technical challenges related to weak, noisy, and non-stationary data analysis for large slow speed bearings. This will be achieved by developing novel algorithms with sparse signal separation, data fusion and machine learning methods, followed by significant demonstration activities on both lab and real world operating environments. The PI has developed the first industrial-scale wind turbine pitch bearing platform including three naturally damaged bearings with over 15 years operating life in a real wind farm and advanced data collection instrument. The newly built platform lays a solid foundation for the proposed research and creates an ideal platform for carrying out demonstration and impact activities. The PI has also secured the unique opportunity to carry out field data collection and demonstration in real world operating wind farms under the strongest supports provided by two industrial project partners.The data collected from three naturally damaged bearings will be made publicly available under open-source licences to enable other researchers to carry out condition assessment for large slow speed bearings. The IP developed during the project will be protected. The developed algorithms will be made publicly available, if not conflicted with the IP.The successful outcome of this project will break new ground in in-situ pitch bearing condition assessment methods and tools, contribute to industrial standards of pitch bearings, and benefit a wide range of industries that use large slow speed bearings, such as offshore oil, gas, mining and steel making, over many decades of bearing service life. The novel methods with regard to weak, noisy and non-stationary data analysis can be used for wide data-driven applications. Therefore, the project has a significant, wide and long term impact in the next few decades.