转炉煤气回收系统优化控制策略应用  

Application of Optimal Control Strategy to Converter Gas Recovery System

作　　者：章家岩[1] 马中海[1] 钱晓斌[2] 李绍铭[1] 郎佳红[1] 

机构地区：[1]安徽工业大学电气信息学院,马鞍山243002 [2]马鞍山钢铁股份有限公司第二轧钢总厂,马鞍山243021

出　　处：《自动化学报》2012年第6期1017-1024,共8页Acta Automatica Sinica

摘　　要：分析了转炉煤气回收工艺特点以及影响回收效果的重要因素,阐述了实现回收过程运行优化控制的相关工艺参数指标,提出了一种基于提高CO浓度的优化控制方案,利用模糊径向基函数(Radical basis function,RBF)神经网络在线辨识出炉口压差与CO浓度之间的数学模型,根据辨识模型实吋调整压差控制同路设定值,通过控制系统跟踪调整后的设定值,在辨识的过程中改进了网络学习算法,使辨识网络对学习参数变化具有较好的鲁棒性,并易于收敛.在应用此优化控制方法对煤气回收系统进行仿真分析的基础上,投入现场应用,结果表明,此优化控制策略能明显提高煤气回收的质量和品质,达到了良好的实际应用效果.The characteristics of converter gas recovery process and the important factor in the effect of recycling are analyzed,run indicators of process parameters to achieve optimal control of the recycling process are described,and an optimal control strategy to increase CO concentration is raised.Fuzzy radical basis function（RBF） neural network is used to online identify the mathematical model between pressure of converter mouth and the CO concentration.The identified model is used to optimize pressure settings and control the pressure of converter mouth near the set value,thus significantly increasing the effect of CO concentration.Network learning algorithm is improved in the process of identification,so that the network is robust and easy to convergence for the variational learning parameters.The application of the optimal control strategy in a gas recovery system shows that this optimal control strategy can significantly improve the quality and quality of gas recovery,and achieve good application results.

关 键 词：煤气回收 炉口压差 模糊径向基函数 CO浓度优化 学习速率 

分 类 号：TF713[冶金工程—钢铁冶金] TP273[自动化与计算机技术—检测技术与自动化装置]