Vol.24 卢虎生等：高炉炉况判断神经网络专家系统 ·279· 于实际生产，取得了令人满意的结果 5刘金琨，王树青.高炉异常炉况神经网络专家系统) 钢铁研究学报，1998,10(3)：67 参考文献 6 Israel Broner,Carlton R Comstock.Combining Expert Sys- 1刘云彩.当代高炉炼铁成就】炼铁，2001,20(3)27 tems and Neural Networks for Learning Site-specific 2毕学工.人工智能和专家系统在钢铁工业中的应用 Conditions[J].Computers and Electronics in Agriculture [).武汉钢铁学院学报，1995,18(2)：146 1997,19:37 3 Lourdes Mattos Brasil,Fernando Mendes de Azevedo, 7 Enbo Feng,Haibin Yang,Ming Rao.Fuzzy Expert System Jorge Muniz Barreto.A Hybrid Expert System for the Di- for Real-time Process Condition Monitoring and Incident agnosis of Epileptic Crisis []Artificial Intelligence in Prevension[J].Expert systems with Applications,1998,15: Medicine,2001,21:227 383 4杨尚宝，杨天钩，董一诚.神经网络高炉炉况预测与 8中国软件行业协会人工智能协会.人工智能辞典M 判断专家系统J.北京科技大学学报，1996,18，(3)：220 北京：人民邮电出版社，1995 Neural Network Expert System of Forecasting Blast Furnace Operational Conditions LU Husheng2 GAO Bin,ZHAO Liguo,GUO Hongwe?,YANG Tianjun 1)Information Engineering School,UST Beijing,Beijing 100083,China 2)Baotou University of Iron and Steel,Baotou 014010,China 2)Metallurgy School,UST Beijing,Beijing 100083,China ABSTRACT Based on deeply-analyzing the characteristics of iron-making process,it is presented that gen- eralization and self-adaptation of the BF judgement systems are two important factors for maintaining the sta- bility and efficiency of neural network expert system.The strategy for improving these two features has been proposed and a new developed system has been proved to be satisfactory in the on-site blast furnace operation. KEY WORDS expert system;neural network;blast furnace;generalization;selp-adaptation 望里ee业ea堂ases堂ee堂堂SPes pesfespe堂业望PesPeSYesRooReeTesYeoTeote堂s (上接第275页) Refining Technology by Inocualting Clean Steel with Titanium Nitride CHENG Guoguang",ZHU Xiaoxia,PENG Yanfeng",WANG Yugang,ZHAO Pep I)Metallurgy School,UST Beijing,Beijing 100083 2)Materials Science and Engineering School,UST Beijing,Beijing 100083,China 3)Central Iron and Steel Research Institute,Beijing 100081,China ABSTRACT Precipitation and nucleation of TiN during solidification of clean steel have been studied.And the possibility of using TiN to refine as-cast grains as heterogeneous nucleation sites as well as to reduce mac- rosegregation in continuous casting of steel is discussed.The conditions that TiN precipitates at the beginning of solidification have been acquired and the effectiveness of this TiN refining as-cast structure process has been investigated with comparative experimental methods.It is shown that refining grains by inoculating clean steel with TiN is an effective way provided that the process is controlled strictly. KEY WORDS TiN:solidification;grain refinement;clean steel;precipitationl V b . 4 卢2 虎生等 : 高炉 炉况判 断神经 网络专 家系 统 . 7 , 2 . 于 实际生产 , 取得 了令人满意 的结果 . 参 考 文 献 1 刘云彩 . 当代高炉 炼铁 成就 [J] . 炼 铁 , 2 0 01 , 2 0 ( 3) : 27 2 毕学 工 . 人 工智能 和专家 系统在 钢铁 工业 中的应用 【J ] . 武汉 钢铁学 院学 报 , 19 9 5 , 18 ( 2 ) : 1 4 6 3 L o ur d e s M at o s B r a s i l , F e nr a n d o M e n d e s d e A ez v e d o , J o gr e M u n i z B a er t o . A yH b r i d E x P e rt S y s et m fo r ht e D i - a gn o s i s o f E P il e Pti e C r i s i s ! J ] . A rt iif e i a l I n t e l li g e n e e i n M e d i e i n e , 2 0 0 1 , 2 1 : 2 2 7 4 杨 尚宝 , 杨 天钧 , 董 一诚 . 神 经 网络 高炉 炉况预测 与 判断专 家 系统 [ J ] . 北京科 技大学 学报 , 1 9 9 6 , 1 8 , ( 3 ) : 2 2 0 5 刘金馄 , 王树青 . 高炉异 常炉况神 经网络专家系统 [J] . 钢铁 研究学 报 , 1 9 9 8 , 1 0 ( 3 ) : 6 7 6 I s aer l B r o n e ’r C ar l t o n R C o m s ot c k . C o m bi n i n g E xP e rt Sy s - t e m s an d N e uar l N e tw o kr s fo r L e am i n g S iet 一 s P e e i if e C o n d i ti o n s [J ] . C o m P ut e r s an d E l e e tr o n i e s i n A gr i e u lut r e , 19 9 7 , 1 9 : 3 7 7 E n b o F e n g , H ia bi n ya n g , M i n g Ra o . F u Z y E xP e rt s y s t e m fo r eR a l 一 t im e P r o e e s s C o n d i t i o n M o n i t o r i n g an d I n e i d e n t P r e v e n s i o n [J ] . E xP e rt s y s t e m s w i th A PP li e at i o n s , 19 9 8 , 1 5 : 3 8 3 8 中国软件行 业协会人 工智能 协会 . 人工智 能辞典 IM I 北 京 : 人 民 邮 电出版社 , 1 9 5 N e ur a l N e wt o r k E xP e rt S y s t e m o f F o r e e a s t i n g B l a s t F um a c e O P e r at i o n a l C o n d it i o n s L U uH s h e 叮 , , , , , GA O B in , ), IZ ￡刁o L恻 。 , ), G Uo OH n g w e尸,) YA N G iT anj u n , , l ) I n fo mr at i o n E n g i n e e r i n g S e h o o l , U S T B e ij i n g , B e ij i n g 10 0 0 8 3 , C h i n a Z ) B ao t o u U n i v e sr iyt o f lor n an d s t e e l , B ao t o u o l 4 0 10 , C h i n a 2 ) M et a ll u r g y S c h o o l , U S T B e ij i n g , B e ij i n g 10 0 0 8 3 , Ch i n a A B S T R A C T B a s e d o n d e e P ly 一 an a ly z i n g t h e e h ar a e t e r i s t i e s o f ior n 一 m a k l n g Por e e s s , it i s rP e s e nt e d th a t g e n - e ar li z at i o n a n d s e l-f a d a Pt at i o n o f ht e B F j u d g e m e in s y s t e m s ar e wt o 1m P o rt a n t fa e t o r s fo r m a i n at i n i n g ht e s t a - b iliyt a n d e if c i e n e y o f n e u r a l n e wt o kr e x P e rt s y s t e m . T h e s t r a t e gy fo r im P r o v i n g ht e s e wt o fe a ot r e s h a s b e e n P r o P o s e d a n d a n e w de v e l o Pe d s y s t e m h a s b e e n Por v e d t o b e s at i s fa e t o yr i n ht e o n 一 s it e b l a st fu m a e e op e r at i o n . K E Y WO R D S e XP e rt s y s t e m : n e ur a l n e wt o kr ; b l a s t ft l rn a e e : g e n e r a li atZ i o n : s e lP 一 a d ap t at i o n (上接第 2 7 5 页 ) R e if n i n g eT e hn o l o g y b y I n o c u a lt i n g C l e an S t e e l w it h iT t a n i um N it r i d e c H万N G uG o g u a gn , ’ , ZH U iX a ox ial ’ , 尸皿叭子aY fen 心 , , 洲刃G h 心a gn , ’ , Z H 刁口 eP 尸 ’ l ) M e t a ll u gr y S e h o o l , U S T B e ij i n g , B e ij i n g 10 0 0 8 3 2 ) M at e r i a l s S e i e n e e an d E n g i n e e ir n g s e h o o l , U S T B e ij i n g , B e ij i n g l 0 0 0 8 3 , Ch i n a 3 ) C e nt r a l I ro n an d S t e e l R e s e are h I n s t itu te , B e ij in g 10 0 0 8 1 , C h i n a A B S T R A C T P r e e iPit at i o n a n d nu e l e at i o n o f T NI d u r l n g s o lid iif e at i o n o f e l e an s t e e l h va e b e e n s t u d i e d . A n d t h e P o s s ib ility o f u s i n g T NI t o re if n e a s 一 e a st gr a i n s a s h e t e r o g e n e o u s nu e l e at i o n s ite s a s w e ll a s t o r e du e e m a e - r o s e gr e g at i o n i n e o nt inu o u s c a s t i n g o f s t e e l 1 5 d i s e u s s e d . T h e e o n d it i o n s ht at TiN Pre e iPiat e s at ht e b e g in i n g o f s o lid iif e at i o n h va e b e e n a c qu ire d an d ht e e fe c t i v e n e s s o f t h i s T iN re if n ign a s 一 e a st s tru e trIJ e P r o e e s s h a s b e e n i n v e s t i g a t e d w iht c o m P ar at i v e e xP e ir m e n t a l m e ht o d s . It 1 5 s h o w n ht at r e fm i n g gr a i n s 勿 i n o c u lat i n g e l e an s t e e l w it h T IN 1 5 an e fe c t i v e w ay P r o v id e d ht at ht e Pr o e e s s 1 5 e o n tr o ll e d s tr i c t ly . K E Y W O R D S T NI : s o lid iif e at i o n ; gr a i n r e if n e m e in : e l e an s t e e l: Pr e e iP iat i o n