使用反问题和遗传算法监测高炉炉底热侵蚀  

Monitoring the Wear-line of Blast Furnace with Inverse Problem and Genetic Algorithm

作　　者：陈怡[1] 谭永基[1] 顾祥林[2] 陈贺林[2] 

机构地区：[1]复旦大学数学研究所,上海200433 [2]上海宝山钢铁股份有限公司,上海201900

出　　处：《复旦学报（自然科学版）》2006年第2期170-176,共7页Journal of Fudan University：Natural Science

摘　　要：把确定高炉炉缸炉底热侵蚀边界归结为稳态热传导方程的反问题,并转化为极小化在若干测温点处由设定侵蚀边界对应的偏微分方程边值问题求得的温度与实测温度之间误差的优化问题.使用经代数交叉、动态变异改进的实数编码的遗传算法,成功实现优化计算.采用模拟数据与真实数据进行的数值试验均表明:计算结果精确可靠,抗干扰性强.The problem of monitoring the wear-line of a blast furnace is treated as the inverse problem of the static heat conduction system, and the latter has been turned into the optimization of the squared residual of the measured and ealculated temperature at the sensor locations. The optimization has been carried out by the Genetic Algorithm adjusted with algebraic cross and dynamic mutation.Both of the numerical experiments of teat model data and industrial data have confinned the accuracy and robust of our algorithm.

关 键 词：高炉 侵蚀 遗传算法 正则化 

分 类 号：TB115[理学—数学]