基于具有明显测量误差的辐射图象的二维炉膛温度分布重建和快速特征识别研究.docxVIP

2000 年 9 月Proceedings of t he CSE 2000 年 9 月 Proceedings of t he CSEE Sep . 2000 文章编号 :025828013 (2000) 0920067205 基于具有明显测量误差的辐射图象的 二维炉膛温度分布重建和快速特征识别研究 韩曙东 , 周怀春 , 盛 锋 , 黄勇理 , 孙学信 (华中理工大学煤燃烧国家重点实验室 ,湖北省 武汉市 430074) STUDY O N RECO NSTRUCTIO N A ND FAST C HA RACTERISTIC RECO GNITIO N OF TEM ERATURE D ISTRIBUTIO NS IN A 22D FURNACE VIA RAD IA NT ENERGY IMA GES WITH SENSITIVE M EAS UREM ENT ERRO RS HAN Shu2do ng , ZHOU Huai2chun , SH EN G Feng , HUAN G Yo ng2li , SU N Xue2xin ( Natio nal L abo rato ry of Coal Co mbustio n , Huazho ng U niversit y of Science and Technology , Wuhan 430074 , China) ABSTRACT: Reconst ruction of temperat ure dist ribution t hrough radiant image processing in 22D f urnaces is a typical inverse prob2 lem , and it is very sensitive to measurement errors by only one CCD camera. Radiant images capt ured by four cameras are used to add more usef ul information , and t he inverse problem is converted into a kind of linear programming problem. Through adding const raint s of t he temperat ure bounds in it s dist ribution , an interior point algo2 rit hm is adopted to solve t he linear programming and forms a met hod of reconst ruction and fast characteristic recognition of tem2 perat ure dist ribution in 22D f urnaces. Calculation result s for single2 peaked and double2peaked temperat ure dist ributions in a 10m ×10m f urnace indicate t hat t his algorit hm can remarkably improve t he ac2 curacy of t he reconst ruction result s wit h sensitive measurement er2 rors , and can satisfy completely t he requirement s for application. This met hod can be easily extended into reconst ruction of 32D tem2 perat ure dist ributions of utility boiler f urnaces and applicatio ns in ot her related fields. 膛温度分布重建和快速特征识别方法 。对单峰和双峰温度 分布的计算结果表明 ,此方法可以显著提高整个系统在辐射 图象具有明显测量误差情况下的重建精度 ,满足实际应用的 要求 。该方法很容易推广到电站锅炉炉膛三维温度场重建 及其它领域的应用中 。 关键词 :二维温度场 ; 辐射图像处理 ; 线性规划 ; 内点法 中图分类号 : T K311 文献标识码 :A 1 引言 燃烧温度测试技术是燃烧理论和技术发展的重 要基础之一 。由于工业炉膛尺寸太大 ,所以到目前 为止还未建 立 起