Remaining useful life and lifetime extension of wind turbine drivetrains

风力涡轮机传动系统的剩余使用寿命和寿命延长

• 批准号: 2505974
• 金额: --
• 依托单位: University of Strathclyde
• 依托单位国家: 英国
• 项目类别: Studentship
• 财政年份: 2020
• 资助国家: 英国
• 起止时间: 2020 至 无数据
• 项目状态: 未结题
• 来源: https://gtr.ukri.org/projects?ref=studentship-2505974
• 关键词: Remaining; useful; life; lifetime; extension

To optimally make decisions for wind turbine maintenance, predictions on the future health states of the wind turbine drivetrain must be carried out. Prognostics is the process whereby past and present condition monitoring data of a system or component is used to estimate its health state into the future. The wind turbine drivetrain is a critical subassembly in terms of downtime and replacement costs; therefore, it is very important to monitor it and perform accurate prognostics. Monitoring is usually done using vibration, SCADA, and oil data. An integrated decision support system fusing the aforementioned multiple sources of data can increase the confidence of a maintenance action under a condition-based monitoring scheme.This EngD will focus on the wind turbine drivetrain fault detection, isolation and remaining useful life esti-mation using advanced signal processing and machine learning methods and considering component de-pendencies. The work will involve the following:*Research of various signal processing methods applicable to wind turbine vibration signals.*Extraction of health indicators from multiple data sources/streams.*Research of various applicable machine learning methods for prediction using the extracted indica-tors as features.*Model dependencies between drivetrain components.*Develop a multi-component degradation model.*Model lifetime extension scenarios.The work will be validated using data from operating wind farms.As a collaborative research project, the research student will work together with Natural Power and the University of Strathclyde research teams, spending time in both organisations.

为了最佳地做出风力涡轮机维护的决策，必须执行对风力涡轮机传动系的未来健康状态的预测。预测是一个过程，其中过去和现在的状态监测数据的系统或组件是用来估计其健康状态到未来。就停机时间和更换成本而言，风力涡轮机传动系统是一个关键的子组件;因此，对其进行监控并执行精确的动力学分析非常重要。通常使用振动、SCADA和油数据进行监测。综合决策支持系统融合了上述多个数据源，可以在基于状态的监测方案下提高维护行动的置信度。该EngD将重点关注风力涡轮机传动系统故障检测、隔离和剩余使用寿命估计，使用先进的信号处理和机器学习方法，并考虑组件依赖性。本论文的主要工作包括：* 研究适用于风力涡轮机振动信号的各种信号处理方法。*从多个数据源/数据流中提取健康指标。研究各种适用的机器学习方法，使用提取的指标作为特征进行预测。*动力传动系统组件之间的模型依赖性。*建立多组分降解模型。模型寿命延长情景。这项工作将使用来自运营风电场的数据进行验证。作为一个合作研究项目，研究生将与Natural Power和斯特拉斯克莱德大学的研究团队一起工作，在两个组织中花费时间。