RBF神经网络在加速冷却控制系统中应用.pdfVIP

第26卷　第2期 自　动　化　学　报 V o l26,N o 2 2000年3月 A CTA AU TOM A T ICA S IN ICA M ar. , 2000 RBF 神经网络在加速冷却控制 系统中的应用1) 柴天佑　王笑波 ( 东北大学自动化研究中心　沈阳　110006) 摘　要　针对中厚钢板加速冷却过程, 钢板入口温度波动大, 钢板终冷温度难以实现闭环控 制等特点, 首先将RBF (R adial Basis Functions) 神经网络与优化设定技术相结合应用到加速 冷却系统中, 提出了加速冷却控制系统中基于RBF 神经网络的每幕流量的控冷模型. 仿真实 验表明: 该控冷模型能够实现对钢板加速冷却过程的有效控制. 关键词　 神经网络　加速冷却　控冷模型 RBF APPL ICAT ION OF RBF NEURAL NETWORKS IN CONTROL SY STEM OF THE SLAB ACCEL ERAT ING COOL ING PROCESS 　 CHA I T ianyou W AN G X iaobo (R esearch Center of A u tom ation , N ortheastern U niversity , S heny ang 　110006) Abstract 　 In the slab accelerating coo ling p rocess, the range of the slab’s inlet , temperature variance is large and clo sed loop contro l of the slab finish ing coo ling temperature is difficult. In th is paper , RBF neural netwo rk s are emp loyed in the , slab accelerating contro l system s and an RBF neural netwo rk based coo ling contro l model fo r each curtain’s rate is p ropo sed. Sim ulation results and industrial experim ents show that effective contro l of the slab accelerating coo ling p rocess is . realized using th is contro l model 　 , , . Key words RBF neu ral netw o rk s accelerating coo ling coo ling con tro l m odel 1　引言 ( 中厚钢板轧制后的加速冷却工艺, 是在精轧机与热矫直机之间增加一段冷却装置 如