摘要
为了提高铜转炉的操作水平,采用操作模式来描述一组需要在线决策的一组操作参数,提出了基于神经网络和带混沌变量的混沌遗传算法的铜转炉生产过程操作模式智能优化方法.首先,从历史样本集中筛选优化的样本;然后采用BP(Back_Propagation)神经网络来学习优化样本集的优化目标与工艺参数的函数关系;最后采用带混沌变量的混沌遗传算法来寻求优化的操作模式.将该方法应用到铜转炉操作参数的实时优化,工业现场运行结果表明,该方法使转炉产量提高了6%,冷料处理量提高了7.8%.
Operational patterns are used to describe a set of on-line operational parameters which need to be determined on line,and an intelligent method based on neural network combined with improved chaotic genetic algorithm for optimal operational patterns is proposed to improve the operational level of copper converting furnace.Firstly,the optimal samples set is filtered from the historical samples set.Then,the functional relation with objective and technical parameters is trained by a neural network model.Finally,chaotic genetic algorithm with chaotic variables is used to search the optimal operational pattern.The running results of an intelligent system of optimal operational parameters based on the method in the process of copper converting furnace show that the output of converter increases by 6%,and the amount of the treated cool materials rises by 7.8%.
出处
《控制理论与应用》
EI
CAS
CSCD
北大核心
2005年第2期243-247,共5页
Control Theory & Applications
基金
国家973计划资助项目(2002cb312200)
国家自然科学基金资助项目(50374079)
教育部科技研究重点项目(02146)
湖南省自然科学基金资助项目(01JJY2110).
关键词
神经网络
遗传算法
铜转炉
操作参数优化
neural network
genetic algorithm
copper converting furnace
operational parameters optimization
