【作者】 张永辉； 邵诚；

【Author】 Zhang Yonghui 1,2,Shao Cheng 3 1.Optomechatronics Key Lab.,Research Institute of Tsinghua University in Shenzhen,Shenzhen 518057,P.R.China 2.College of Information Science & Technology,Hainan University,Haikou 570228,P.R.China 3.Institute of Advanced Control Technology,Dalian University of Technology,Dalian 116024,P.R.China

【机构】 深圳清华大学研究院光机电重点实验室； 海南大学信息科学与技术学院； 大连理工大学先进控制研究所；

【摘要】 铝粉氮气雾化和分级工艺过程具有强烈的非线性、强耦合、大滞后、不确定性等特点,很难用单一的任何一种智能化方法或传统方法来实现雾化过程中的全局优化控制问题。本文对铝粉氮气雾化过程的工艺特点和控制要求进行了分析和研究,在对铝粉氮气雾化过程的机理深入分析的基础上,提出利用集成优化控制技术来实现铝粉氮气雾化过程的实时有效的控制和过程优化,实现了对铝粉雾化炉温度、循环氮气的稳定控制、利用RBF神经网络建立了铝粉氮气雾化过程模型,以优化铝粉粒度分布为生产目标,采用改进的遗传算法实现了雾化过程优化控制,显著地改善了铝粉雾化效果,有效地提高了球型微细铝粉收率。更多还原

【Abstract】 The process of aluminum powder nitrogen atomizing has characteristics of strong nonlinear,strong coupling,large lag,and uncertainty and so on.It is difficult to achieve global optimal control for atomizing process by using single intelligent or conventional method.In this paper,research work to the techniques features and control demands of the aluminum powder nitrogen atomizing process is done,and integrated optimal control technology is presented to effectively control the aluminum powder nitrogen atomizing process and optimize it.As a result,the temperature of atomizing furnace is stabilizing for melted aluminum atomizing;the concentration of oxygen,pressure and temperature of the recycle nitrogen are all controlled very well;the aluminum powder nitrogen atomization process model based on RBF Neural Networks is presented and enhanced GA based aluminum powder nitrogen atomizing process integrated optimal control is implemented to improve the atomizing effect of aluminum powder and promote the percentage of super-tiny aluminum powder greatly.更多还原