基于改进BOF算法的压板状态识别方法研究

随着电网规模的扩张以及二次技术的发展，二次屏柜的压板数量急剧上升。二次压板一旦出现保护误动或拒动，将严重威胁到电网、设备与人身安全。设计了一种基于改进图像检索算法的继电保护压板状态识别系统对压板退投状态进行识别。首先，对压板图像利用背景差分法及光流法进行前景提取；然后，采用图片增强处理、边缘检测等一系列技术识别开关的投退状态；最后，再将获得的开关状态与系统数据库对比，生成可视化报表，由此判断出压板状态是否正确。提出的方法能够快速识别压板状态正确与否，具有较高识别率，极大缩减了二次操作后核查及日常巡视的时间。

Condition Recognition System for Relay Protection Plate Based on Improved Bag of Feature Algorithm

With the expansion of power grid scale and the development of second technology， the number of press plates of secondary panel has risen sharply. The mal-operation and refusal-operation of the secondary press plate will seriously threaten the safety of power grid， equipment and personnel. This paper designs a condition recognition system for relay protection plate based on an improved image retrieval algorithm to identify the on/off state of the press plate. Firstly，background difference method and optical flow method is used to extract the foreground of the plate image. Then a series of technologies such as image enhancement processing， edge detection and so on are used to identify the switching state of the switch. Finally， the obtained switch state is compared with the system database， and a visual report is generated to determine whether the state of the plate is correct. The proposed system can quickly identify whether the state of the plate is correct or not， has a high recognition rate， and greatly reduces the time required for verification and daily inspection after secondary actions.