基于智能图像处理和三维建模技术的特高压换流阀厅用套管运行状态监测方法  

Operation Condition Monitoring Method of Bushing for UHV Converter Valve Hall Based on Intelligent Image Processing and 3D Modeling Technology

作　　者：张施令 何永胜 宫林 郭强[3] ZHANG Shiling;HE Yongsheng;GONG Lin;GUO Qiang(State Grid Chongqing Electric Power Company Chongqing Electric Power Research Institute,Chongqing 401123,China;State Grid Chongqing Electric Power Company,Chongqing 400014,China;Chongqing University of Technology,Chongqing 400054,China)

机构地区：[1]国网重庆市电力公司电力科学研究院,重庆401123 [2]国网重庆市电力公司,重庆400014 [3]重庆理工大学,重庆400054

出　　处：《绝缘材料》2022年第1期87-94,共8页Insulating Materials

摘　　要：从智能图像处理和三维造型技术的角度提出一种针对特高压换流阀厅用套管运行状态特征的监测方法。该方法主要包括:应用智能图像处理技术对红外热像仪、紫外成像仪数据库进行辨识、分类处理;通过Kalman滤波技术针对典型金具绝缘距离进行实时在线测量;应用基于有限元法三维造型技术建立阀厅典型主设备套管的电场模拟模型,获得关键金具表面电场分布情况。结合图像数据库信息、绝缘距离信息和典型主设备套管电场分布信息有效获取其运行状态参量,并应用智能算法对其运行状态进行自动评估。结果表明:未使用Kalman滤波技术时,在0~160 s内对于关键位置处的距离预估偏差跳变较为剧烈,使用Kalman滤波技术后预估偏差跳变幅度缩小,基本控制在同一误差水平;神经网络具有较好的学习效果,适应度函数迭代100次后趋于稳定,且模糊神经网络局部权值出现较为典型的非线性特征。研究结论可有效发掘潜伏性隐患、定位正发性故障,为主设备套管运行维护提供有效的数据支撑和保护策略。A monitoring method for the operating state characteristics of bushing in UHV converter valve hall was proposed from the perspective of intelligent image processing and three-dimensional modeling technology.This method mainly included using intelligent image processing technology to identify and classify the databases of infrared thermal imager and ultraviolet imager,using Kalman filtering technology to realize on-line measurement in real time for the insulation distance of typical metal fittings,and using three-dimensional modeling technology based on finite element method to build the electric field simulation model of typical main equipment bushing in valve hall,and the electric field distribution on the surface of key metal fittings was obtained.Combined with the image database information,insulation distance information,and electric field distribution information of typical main equipment bushing,its operation state parameters were obtained effectively,and its operation state was evaluated by the intelligent algorithm automatically.The results show that when the Kalman filtering technology is not used,the prediction deviation of the distance at key positions changes severely in 0‒160 s.After using the Kalman filtering technology,the jump amplitude of prediction deviation decreases and is basically controlled at the same error level.The neural network has good learning effect,the fitness function tends to be stable after 100 iterations,and the local weight of fuzzy neural network shows typical nonlinear characteristics.The research conclusion can explore potential hidden dangers and locate positive faults effectively,which provide effective data support and protection strategies for the operation and maintenance of main equipment bushing.

关 键 词：特高压换流阀厅 套管 智能图像处理 三维造型技术 Kalman滤波 

分 类 号：TM216[一般工业技术—材料科学与工程]