电力设备缺陷智能识别系统设计  

Design of Intelligent Recognition System for Defects of Power Equipment

作　　者：曹熙 CAO Xi(Digital Grid Research Institute,China Southern Power Grid.,Guangzhou Guangdong 510000,China)

机构地区：[1]南方电网大数据服务有限公司,广东广州510000

出　　处：《信息与电脑》2022年第2期93-95,共3页Information & Computer

摘　　要：针对传统电力设备缺陷智能识别系统信号处理能力较差、易出现识别结果准确度较低的问题,设计了新型电力设备缺陷智能识别系统。首先,选用STM32F107信号控制器对信号采集过程进行控制,并将中央控制器芯片设定为STM32F107保证系统的运行效果;其次,使用卷积神经网络构建信息分类器,对电力系统设备信号信息进行分类处理,将此分类结果作为识别的数据来源;最后,在原有的卷积网络中增加归一处理层,对设备的缺陷信息进行分类,完成设备缺陷识别。经系统功能与性能测试结果融合分析可知,设计系统基础性能较为良好,可有效控制识别准确性。Aiming at the problem of poor signal processing ability and low accuracy of identification results of traditional power equipment defect intelligent identification system, a new power equipment defect intelligent identification system is designed. Firstly,stm32 f107 signal controller is selected to control the signal acquisition process, and the central controller chip is set as stm32 f107 to ensure the operation effect of the system;Secondly, the convolution neural network is used to construct an information classifier to classify the signal information of power system equipment, and the classification results are used as the data source of recognition;Finally, a normalization processing layer is added to the original convolution network to classify the equipment defect information and complete the equipment defect identification. The fusion analysis of system function and performance test results shows that the basic performance of the designed system is relatively good, which can effectively control the recognition accuracy.

关 键 词：电力设备 缺陷识别 Openstack云计算 设备状态检修 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]