基于粗糙集的公用变压器的负荷监测方法  

Load Monitoring Method of Public Transformer Based on Rough Set

作　　者：邰刚 侯玉 TAI Gang;HOU Yu(Shenzhen Power Supply Bureau Co.,Ltd.,Shenzhen 518001,China)

机构地区：[1]深圳供电局有限公司,广东深圳518001

出　　处：《电工技术》2023年第17期25-27,共3页Electric Engineering

摘　　要：负载率是变压器运行工况的重要指标,但在负荷监测过程中,易受后台数据量影响,从而会降低监测精准度,并增加维修的人工成本,浪费大量的耗材和时间。基于此,提出一种基于粗糙集的负荷监测模型。首先对变压器负荷信号进行粗糙集处理,减少后台预处理数据量,缩短负荷预警时间。然后对负荷梯度、负荷邻域进行差值分析,并赋予监测权重。最后,利用粗糙集邻域搜索法进行变压器负荷异常信号识别,并输出最终变压器负荷结果。研究结果显示,粗糙集邻域搜索法能快速、准确地监测变压器负荷信息,监测结果受后台数据量的影响小于10%,监测准确率和监测指标完整率大于90%,优于传统的人工评估方法,满足变压器负荷监测要求,适合变压器负荷预警分析需要。Load rate is an important indicator of transformer operation condition,but in the process of load monitoring,the influence of the amount of background data easily causes reduction of monitoring accuracy heightened the labor cost of maintenance,and remarkable waste of supplies and time.Therefore,a load monitoring model based on rough set is proposed in this paper.Firstly,the transformer load signal is treated by rough set to reduce the amount of background preprocessing data and shorten the load warning time.Then the load gradient and load neighborhood are analyzed,and the monitoring weight is given.Finally,the rough set neighborhood search method is used to identify the abnormal signal of transformer load and output the final transformer load result.The results show that the proposed rough set neighborhood search method can quickly and accurately monitor the transformer load information,and can obtain monitoring results better than the traditional manual evaluation method,as the influence of background data is less than 10%,and the monitoring accuracy rate and the monitoring index integrity rate are higher than 90%.Therefore,the method can meet the requirements for transformer load monitoring and is suitable for early warning analysis of transformer load.

关 键 词：负荷监测 粗糙集 邻域搜索 变压器 

分 类 号：TM406[电气工程—电器]