燃煤锅炉屏式过热器结渣可视化在线监测技术应用研究OA
Research on the Application of Visual Online Monitoring Technology for Slagging in the Platen Superheater of Coal-Fired Boilers
为了实现锅炉屏式过热器结渣情况的可视化在线监测,设计了一种可将镜头伸入炉内拍摄的炉内图像采集系统,在锅炉热态运行时定时定点自动拍摄屏式过热器底部画面,并基于YOLOv8模型改进的YO-LO-LTC图像识别算法对拍摄到的屏式过热器结渣视频进行识别训练.结果表明:炉内图像采集系统可连续稳定工作,减少人工观测的工作量;改进后的YOLO-LTC算法可以有效识别出不同时间屏式过热器的结渣比例,并实时显示到分散控制系统页面,实现了屏式过热器结渣在线监测,识别结果中严重结渣样本数仅占4.20%,且日期分布较散,说明屏式过热器吹灰频次过高,后续可适当降低频次.
To achieve visual online monitoring of slagging in the platen superheaters,a system was designed for capturing in-furnace images with a camera that can be extended into the furnace.The system automatically captures images of the bottom of the platen superheater at set intervals during the boiler's hot-state operation.Using the YOLOv8-based improved YOLO-LTC image recognition algorithm,the captured video of the platen superheater slagging was subjected to recognition and training.The results show that the in-furnace image acquisition system operates continuously and stably,reducing the workload of manual observation.The improved YOLO-LTC algorithm can effectively identify the slagging proportions of the platen superheater at different times and display them in real time on the DCS system page,realizing the online monitoring of slagging in the platen superheater.In the recognition results,the number of severely slagged samples only accounts for 4.20%,and the dates are rather scattered,indicating that the current soot blowing frequency for the platen superheater is too high and it can be appropriately reduced in the future.
杨倩;王晓涛;温小豪;刘逸;李培
陕西榆林能源集团杨伙盘煤电有限公司,陕西榆林 719316陕西榆林能源集团杨伙盘煤电有限公司,陕西榆林 719316浙江大学热能工程研究所能源高效清洁利用全国重点实验室,浙江 杭州 310027陕西榆林能源集团杨伙盘煤电有限公司,陕西榆林 719316浙江大学热能工程研究所能源高效清洁利用全国重点实验室,浙江 杭州 310027
能源科技
屏式过热器结渣可视化图像识别在线监测YOLOv8
platen superheaterslagging visualizationimage recognitiononline monitoringYOLOv8
《锅炉技术》 2026 (4)
62-67,6
评论