基于监测数据的风力发电机故障预警研究

发布时间：2018-11-18 10:24

【摘要】：近年来,随着环境污染和能源危机的日益加剧,风能等可再生清洁能源因具有储量丰富和无污染等特点而备受世界各国的青睐和重视。我国风电行业发展迅速,风力发电机作为一种旋转的机械设备,所具有的零件较多,结构也相对复杂;同时,风电机组工作于人烟稀少,自然条件恶劣的环境中,导致风机在运行过程中故障频繁发生,频繁的故障维修致使风电场的运营成本提高。因此,如何利用智能监控手段减少风机故障次数以达到节约风电场的运营成本的目的,是目前大部分风电场亟需要解决的重要课题。基于此背景下,开展智能手段的风机故障预警及远程监控研究具有重大意义。本文首先研究国内外专家学者针对风力发电机的故障预警和故障诊断的研究现状;其次分析了风机的工作原理、组成部分及典型故障,总结了故障发生的原因;最后,利用数据挖掘技术中的相关性提取相关规则,并将相关规则保存在数据库,通过查询功能实现故障的匹配。重点针对风力发电机故障预警进行了研究,通过采集SCADA系统的监控数据,计算了数据之间的相关系数,分析了影响风力发电机温度的相关参量,并组建相关变量集。在此基础上,建立了基于风力发电机温度的故障预警模型。通过预测值与实际值的残差分析验证了风机温度故障预警模型的有效性。本文结合辽宁龙源风力发电有限公司法库风电场的实际情况,简要说明风机远程监控的设计思路,并详细说明如何实现此方法,其中包括数据挖掘、数据采集、数据传输方法、数据存储、数据发布和监控,以及整个通讯的框架,最终实现在用户的手机APP上监控风电场运行并实施安全保护。对设计其他的监控系统起到了指导作用,同时为以后开发相关软件奠定了基础。
[Abstract]:In recent years, with the worsening of environmental pollution and energy crisis, renewable clean energy, such as wind energy, has attracted much attention from all over the world because of its rich reserves and no pollution. Wind turbine is developing rapidly in our country. As a kind of rotating mechanical equipment, wind turbine has more parts and more complicated structure. At the same time, the wind turbine works in the environment with few people and bad natural conditions, which leads to frequent faults and frequent maintenance of the wind farm. Therefore, how to use intelligent monitoring means to reduce the number of fan failures to achieve the purpose of saving the operating costs of wind farms is an important issue that most wind farms need to solve. Based on this background, it is of great significance to carry out the research of fan fault warning and remote monitoring by intelligent means. This paper firstly studies the research status of wind turbine fault early warning and fault diagnosis at home and abroad, then analyzes the working principle, components and typical faults of the fan, and summarizes the causes of the fault. Finally, the correlation in data mining technology is used to extract the relevant rules, and the relevant rules are saved in the database, and the fault matching is realized by the query function. By collecting the monitoring data of the SCADA system, the correlation coefficient between the data is calculated, the related parameters affecting the temperature of the wind turbine are analyzed, and the relevant variable sets are set up. On this basis, a fault warning model based on wind turbine temperature is established. The validity of the early warning model of fan temperature fault is verified by the residual analysis of the predicted value and the actual value. Based on the actual situation of Faku wind farm in Liaoning Longyuan Wind Power Co., Ltd, this paper briefly explains the design idea of remote monitoring fan, and explains in detail how to realize this method, including data mining, data acquisition, data transmission, etc. Data storage, data release and monitoring, as well as the framework of the entire communication, the final implementation of the user's mobile phone APP to monitor the operation of wind farms and the implementation of security protection. It plays a guiding role in the design of other monitoring systems and lays a foundation for the development of related software in the future.
【学位授予单位】：沈阳工程学院
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TM315

【参考文献】