一种关口电能计量装置智能故障诊断及预警技术

Research on intelligent fault diagnosis and early warning technology of gateway electrical energy metering device

作为电力应用中最重要的日常管理工作,电能计量贯穿于电能生产、传输、使用的全过程,确保电能计量准确可靠地运行,是建立公平、公正、有序的电力营销市场的关键。针对目前关口电能计量装置检验过程中存在工作量大、校验时间长、采样精度和算法性能制约等问题,对电能计量装置远程在线状态诊断技术进行研究,基于大量的电能计量采样数据,引入大数据挖掘中的并行计算,同时采用支持向量机,建立数据和任务并行化故障诊断模型,实现对电能计量装置运行异常特征、故障状态的在线实时监测和故障诊断。利用可视化方式及时上传装置故障图片,同时采用多样化的故障预警技术,以便工作人员分析处理故障或异常状态下的电能计量装置,进而提升关口电能计量装置的远程状态监测和运维管理水平。

As the most important daily management work in power application, electrical energy metering is carried out throughout the whole process of power production, transmission and use. Therefore,ensuring the accurate and reliable operation of electric energy measurement is the key to establish a fair, just and orderly electricity market. At present, there are many problems in the inspection process of electric power measurement devices, such as large workload, long verification time, sampling accuracy and algorithm performance constraints. In this paper, the remote on-line diagnosis of electrical energy metering device is studied. Based on a large amount of sampling data of power measurement, voltage, and current, parallel computing in data mining is presented. Based on the support vector machine,data and task parallelization fault diagnosis models are established while the on-line real-time monitoring and failure warning for electric power measurement device is realized, which contains abnormal characteristics and fault state. The visualization method is used to upload the device failure pictures in time. Meanwhile, the diversified fault early warning technology is adopted, so that the staff can analyze the power metering device under the fault or abnormal state, thereby improving the remote state monitoring and operation and maintenance management level of the electrical energy metering device.