基于光谱分析和支持向量机算法的转炉炼钢终点碳分类检测技术-光学工程专业论文.docxVIP

研究生优秀毕业论文 Abstract Abstract 硕士学位论文 Abstract ne endpoint con仃01 of c曲on content is a key process in tlle production of BOF steelmakiIlg，which is directly related to the quality of steel，steelmakng costs and exhaust emissions．However，in t11e smelting process?due to t11e ins协ili够of the raw material is added and complex chemical reactions in me process of blo埘ng，也e endpoim accurate con订01 of carbon and temperature has been a worldwide problem be s01ved for maIly years in metallurgicaJ industry．Therefore，廿le study of a hi曲 precision detection metllod of 0111ine a11d real一time is particularly urgent．Aiming at t11is problem，a real-time class怕ca矗on detection method of carbon coment of steelis 咖died． TKs p印er aIlalyzes t11e research of BOF endpoint con订ol techn0109y at this Stage，desc而bes some cuHent detection methods．at me s锄e time，slunmarizes恤eir pros and cons．Tms metllod is baSed on the spectmscopy of n锄e at fumace mouth aIld the principle of support Vector maclline．Firstly，thjs paper analyzes t11e spectrm of fl锄e and extracts characteristic v撕a_bles．Secondlv’based on t11e SVM classification algorimm． characteristic v撕ables are trajned， aJld classification model can be obtained．FiIlally，廿1is paper designs real-time detection algori也m of Steelmahng process，t11e exact value of carbon is detected by existing fitting curves，and based on 也e LabVIEW platfoⅡr1．deVelop a real—time detection system at．[11e scene． This method is proVed to be 11igh-precision，real一time detection and s廿’ong a11ti_j煳iIlg abili吼etc．Entirely suitable for site enViroI吼em，and meet廿le practical reauirements of steel． Key words：BOF’endpoint control，Spec订oscopy’support Vector machine，model trajlling． II 硕士学位论文 硕士学位论文 基于光谱分析和支持向量机算法的转炉炼钢终点碳分类检测技术 目 录 摘 要 ．I Abstract ． II 1绪论 。1 1．1课题研究背景及意义 1 1．2转炉炼钢的工艺 l 1．3转炉炼钢终点控制 3 1．4