摘要
随着电力尖峰负荷的逐年攀升,变电站设备温度超高的情况日益严重。为解决变电站设备温度监测难度大的问题,提出了一种基于局部不变特征的变电站设备温度监测方法。首先,通过巡检机器人采集变电站设备的红外图像数据。然后,采用局部不变特征算法提取变电站设备温度区域数据,并进行异常图像区域分割。在此基础上,通过改进热力平均温差法对变电站设备异常图像区域进行校验。最后,在某220 kV变电站应用所提方法,平均监测准确性为99.3%。该结果验证了该方法的有效性。
As the power spike load rises year by year, the situation of super high temperature of substation equipment becomes more and more serious. To solve the problem of difficult substation equipment temperature monitoring, a substation equipment temperature monitoring method based on local invariant characteristics is proposed. Firstly, the infrared image data of substation equipment is collected by the inspection robot. Then, the local invariant feature algorithm is used to extract the temperature region data of substation equipment and into the abnormal image region segmentation. On this basis, the abnormal image region of substation equipment is calibrated by improving the thermal average temperature difference method. Finally, the proposed method is applied in a 220 kV substation with an average monitoring accuracy of 99.3%. This result verifies the effectiveness of the method.
作者
贾东明
JIA Dongming(State Grid Beijing Maintenance Company,Beijing 100073,China)
出处
《自动化仪表》
CAS
2022年第9期14-17,共4页
Process Automation Instrumentation
关键词
局部不变特征
变电站
设备温度
投射校准
图像分割
热力平均温差
Local invariant characteristics
Substation
Equipment temperature
Projection calibration
Image segmentation
Thermal mean temperature difference