基于图像智能识别的凝结水精处理控制装置设计  

Design of Control Device for Condensate Concentrate Treatment Based on Image Intelligent Recognition

作　　者：程胜 CHENG Sheng(National Energy Group Yongzhou Power Generation Co.,Ltd.Yongzhou 425900,China)

机构地区：[1]国家能源集团永州发电有限公司,湖南永州425900

出　　处：《全面腐蚀控制》2025年第3期167-171,共5页Total Corrosion Control

摘　　要：在凝结水精处理系统中,存在失效树脂分离与输送不彻底的问题,为提高树脂再生度,本论文设计了改进的Zernike矩的树脂分界面识别算法,基于自动阈值计算和亚像素求解投射关系,大幅度提高计算树脂体积的准确度。这一技术有效避免了分离过程中阴阳树脂配比混乱的现象,使阴阳树脂分离度显著提高,大幅提高了再生过程的控制精度。该图像智能识别技术与算法在实际生产过程中,增大了系统周期制水量,延长了运行周期,为实现“零排放”目标提供了技术支持。In the condensate concentrate treatment system,there is the problem of incomplete separation and delivery of failure resin.In order to improve the resin regeneration,this paper designs an improved Zernike moment resin interface identification algorithm,based on automatic threshold calculation and sub-pixel solution projection relationship,greatly improve the accuracy of the calculation of resin volume.This technology effectively avoids the phenomenon of chaotic ratio of Yin and Yang resin in the separation process,significantly improves the separation degree of Yin and Yang resin,and greatly improves the control accuracy of the regeneration process.In the actual production process,the image intelligent recognition technology and algorithm increase the periodic water quantity of the system,extend the operation cycle,and provide technical support for realizing the goal of“zero emission”.

关 键 词：树脂输送图像 智能识别 树脂分离与输送 树脂再生 

分 类 号：TM621[电气工程—电力系统及自动化]