基于双通道图像的电力设备智能监测技术研究

Research on intelligent monitoring technology of power equipment based on dual channel image

以目前变电站电力设备故障监测为背景,以变电站绝缘子为对象,利用机器视觉技术,重点开展了基于双通道的图像采集技术、电力设备目标识别以及红外感兴趣区域(ROI)自动定位测温技术的研究。通过硬件设计获取双通道(可见光、红外)图像信息,建立了绝缘子数据库,利用Mask R-CNN算法对绝缘子进行有效识别,利用红外测温技术对红外ROI区域进行自动测温,搭建了一套基于双通道图像的电力设备智能监测系统。实验结果表明:本文提出的算法可以实现精确的绝缘子目标识别,为电力设备的智能监测提供了可行的视觉监测技术途径。

Based on the current substation power equipment fault monitoring,this paper takes the substation insulator as the object,uses machine vision technology to focus on the dual-channel image acquisition technology,power equipment target recognition and infrared interest area(ROI)automatic positioning measurement.The two-channel image information is obtained through hardware design,the insulator database is established,the insulator is effectively identified by Mask R-CNN algorithm,and the ROI area is automatically measured by infrared temperature measurement technology.A set of power equipment based on two-channel image is built.Intelligent monitoring system.The experimental results show that the proposed algorithm can achieve accurate insulator target recognition and provide a feasible visual monitoring technology for intelligent monitoring of power equipment.