大数据环境下智能电网关键设备健康评估

发布时间：2018-04-19 09:18

本文选题：智能电网 + 健康评估　；参考：《华北电力大学》2017年硕士论文

【摘要】：伴随着数字信息化时代的快速发展,信息量也呈爆炸性增长态势。当前信息通信技术与电力生产深度融合,对电力工业的价值贡献已经从量变转变到质变,其最鲜明的体现就是电力数据成为电力工业的核心资产。目前中国的电力系统已成为世界上最大规模的关系国计民生的电力网络。电力设备的可靠性、高效运行与有效管理对电力系统的安全、稳定变得愈来愈重要。如何从海量的电力设备监测数据中快速挖掘和发现设备的健康状态与缺陷信息,成为研究者和电力企业的重要关注点。智能电网中的众多传感器会实时地产生大量数据流,对新型流式数据的分析与处理,给设备的健康评估带来了很大的挑战。数据聚类方法是数据挖掘中的一种重要的数据处理技术,许多研究人员提出了众多具有代表性的聚类算法。然而,新型流式数据的出现使得这些经典聚类算法不能直接运用,故而需要研究新的数据流分析与处理方法。云模型是将随机性与模糊性相结合,通过特定的算法实现定性、定量间不确定转换的一种模型。目前,该模型也受到众多研究者关注,并成功应用于许多领域。基于以上问题本文探讨了一种基于新型数据流聚类方法和云模型的设备健康评估方法。该方法包括离线处理和在线实时处理两个模块。离线处理模块,首先基于设备的正常状态的历史运行数据,运用聚类方法实现设备运行工况空间的划分,并计算每种工况下的设备标准状态组合高斯云;在线实时处理模块,采用流式聚类算法对智能电网设备的实时数据流进行工况辨识,并针对每个聚簇采用微簇的方法获取当前数据流的摘要信息,计算设备实时状态的组合高斯云;之后计算实时状态的组合高斯云与标准状态组合高斯云的偏离值并将其作为设备的健康指数;最后根据健康指数的大小对设备的健康状态进行分级。文末通过实例验证分析,利用风电机组的实时数据流,就本文所探讨的方法进行该风电机组的健康评估。实验结果说明该方法所得出的结论符合风电机组的实际运行情况,并能够对风电机组的健康恶化趋势进行预警。
[Abstract]:With the rapid development of digital information age, the amount of information is also explosive growth trend.At present, the deep integration of information and communication technology and power production has changed the value contribution of power industry from quantitative change to qualitative change, the most obvious manifestation of which is that power data has become the core asset of power industry.At present, China's power system has become the world's largest power network related to the national economy and people's livelihood.The reliability, efficient operation and effective management of power equipment are becoming more and more important for the safety and stability of power system.How to quickly mine and discover the health status and defect information from massive monitoring data of power equipment has become an important concern of researchers and power enterprises.A large number of sensors in smart grid can generate a large number of data streams in real time. The analysis and processing of new flow data brings great challenges to the health assessment of equipment.Data clustering is an important data processing technology in data mining. Many researchers have proposed many representative clustering algorithms.However, due to the emergence of new flow data, these classical clustering algorithms can not be used directly, so it is necessary to study new data flow analysis and processing methods.Cloud model is a kind of model which combines randomness with fuzziness and realizes qualitative and quantitative uncertainty conversion through specific algorithms.At present, the model has been concerned by many researchers, and has been successfully applied in many fields.Based on the above problems, a new method of equipment health assessment based on new data stream clustering and cloud model is discussed.The method includes two modules: offline processing and online real-time processing.Off-line processing module, first of all, based on the normal state of the equipment historical operation data, the use of clustering method to achieve the division of equipment operating conditions space, and calculate the standard state of equipment under each condition of the combination of Gao Si cloud; online real-time processing module,The flow clustering algorithm is used to identify the real time data flow of smart grid equipment. For each cluster, the summary information of the current data flow is obtained, and the combined Gao Si cloud of the real time state of the equipment is calculated.Then the deviations of the combination Gao Si cloud in real time state and Gao Si cloud in standard state are calculated and used as the health index of the equipment. Finally, the health status of the equipment is classified according to the size of the health index.At the end of the paper, the method discussed in this paper is used to evaluate the health of the wind turbine unit by using the real time data stream of the wind turbine through an example verification and analysis.The experimental results show that the conclusions obtained by this method are in line with the actual operating conditions of wind turbines and can be used to warn the deterioration trend of wind turbines' health.
【学位授予单位】：华北电力大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TM76

【参考文献】