摘要
鞍钢4号高炉开发的热状态专家系统，以专家知识为基础判断炉热状态及演变趋势，并将自适应模型、非线性模型和神经元网络模型等三个含硅预报模型结合起来，对热状态、变化趋势及铁水含硅量进行综合预报和推断。运行结果表明：专家系统对高炉向凉、向热预报命中率已超过90％，对铁水含硅量预报命中率在84．7％（±0．1％）。
Based on human expert knowledge the thermal state expert system developed for AISC No 4 BF judges thermal state in blast furnace and its evolving tendency, combines the self-adaptation model,nonlinear model and neural netword modelthree predicting models for Si content and as a result makes a comprehensive predictionand draws an inference as to thermal state,evolving tendency and Si content of hotmetal.The operation resuIt shows that the prediction hitting ratio of the expert systemas to tendency to co1d and tendency to hot has already exceeded 90%,the hitting ratioof hot metal Si-conterit is 84. 7% (±0. 1 % ).
出处
《鞍钢技术》
CAS
1997年第8期6-11,共6页
Angang Technology
关键词
高炉
热状态
专家系统
炉况判断
炼铁
thermal state of blast furnace
Si-content prediction expert system
