摘要
以某钢厂连铸板坯二冷动态控制系统为研究对象,运用神经网络、模糊控制,结合改进PSO算法,设计了动态目标温度控制器、动态表面温度控制器。以此为基础,提出了二冷动态在线学习模糊神经网络控制系统,在遵守冶金准则的前提下,能够根据钢种、拉速、过热度的变化,动态设定目标温度及二冷水量,使二冷前段铸坯温度梯度小,矫直段温度在6℃范围内波动,保证了铸坯的内部及表面质量,降低了二冷用水量。
Dynamic target temperature fuzzy neural network controller and dynamic temperature neural network controller were designed for a introduced secondary cooling dynamic control system in casting slab from a factory based on the neural network,fuzzy control method combining with PSO algorithm.Based on above results,a secondary cooling water dynamic online learning fuzzy neural network control system(LFNC)has been developed which can dynamically control target temperature and secondary cooling water according to variation of steel types,pulling velocity and super-heating degree under the condition of meeting metallurgical criteria to obtain the low temperature gradient at the front stage of secondary cooling and to fluctuate in range of 6 ℃ at straightening stage,improving the quality of the slab and reducing the secondary cooling water.The intelligent control system has the advantages of small calculation and good stability.
出处
《特种铸造及有色合金》
CAS
CSCD
北大核心
2008年第6期463-466,共4页
Special Casting & Nonferrous Alloys
关键词
板坯连铸
二冷
神经网络
模糊
控制
Continuous Slab Casting,Secondary Cooling,Neural Network,Fuzzy,Control