基于图像识别的垃圾焚烧炉膛燃烧状态诊断方法  

Diagnosis Method of Combustion State in Waste Incinerator Based on Image Recognition

作　　者：郭天宇 姚心[2] 贺迪龙 刘兵 GUO Tian-yu;YAO Xin;HE Di-long;LIU Bing

机构地区：[1]中国恩菲工程技术有限公司数字智能技术公司数字化业务部,北京100038 [2]中国恩菲工程技术有限公司,北京100038

出　　处：《有色设备》2022年第6期43-47,共5页Nonferrous Metallurgical Equipment

摘　　要：目前我国城镇生活垃圾的主要处理方式是焚烧处理。国内垃圾水分高、热值低,在焚烧过程中需操作人员时刻观察焚烧炉炉膛的燃烧状态,继而不断调整焚烧炉运行参数,确保焚烧正常稳定进行。人工判断具有很强的主观性,并需大量人力。本文介绍了一种基于图像识别的垃圾焚烧炉膛燃烧状态诊断方法,通过图像处理及人工智能技术,实现对焚烧炉膛垃圾燃烧状态的快速判断,将识别出的火焰燃烧状态接入自动控制系统,可提高垃圾焚烧炉的智能化控制水平。经工程实践验证,采用该方法对火焰状态进行诊断,准确率可达95%。At present, the main treatment method of urban domestic waste in China is incineration. Domestic waste in China has high moisture content and low calorific value. During the incineration process, operators need to observe the combustion state of the incinerator at all times and then constantly adjust the operating parameters of the incinerator to ensure normal and stable incineration. Manual judgment is highly subjective and requires a lot of manpower. This paper introduces a diagnosis method of combustion state in waste incinerator based on image recognition. Through image processing and artificial intelligence technology, we can quickly judge the waste combustion state in the incinerator, and connect the recognized flame combustion state to the automatic control system, which can improve the intelligent control level of the waste incinerator. The engineering practice shows that through this method of combustion state diagnosis, the accuracy can reach over 95%.

关 键 词：图像识别 状态诊断 垃圾焚烧 

分 类 号：X705[环境科学与工程—环境工程]