采用基于改进BP算法的人工神经网络模型，提前预报了烧结矿的R和TFe、SiO2含量，将模型的预报结果转化为规则的输入，设计了基于经验规则的专家系统，结合R、TFe的变化趋势和配料计算提前调整原料的配比。系统正式投入运行后，烧结矿碱度（R）预报命中率达到91%，全铁（TFe）预报命中率达到94%，操作指导建议采纳率为92％，实现了对烧结矿化学成分的稳定控制。
An artificial neural network model based on modified BP algorithm is used to predict R, TFe and SiO2 contents of sinter. The results of predictive model are transformed to input of rules. An expert system based on empirical rules was designed. Combining the change trend of R, TFe with burden calculation, the burden is adjusted with such system. The hitrate of sinter basicity was 91%, that of TFe was 94%, and the acceptance rate of operation suggestions was 92%, realizing the goal of controlling sinter chemical composition steadily.
参考文献
上一张
下一张
上一张
下一张
计量
- 下载量()
- 访问量()
文章评分
- 您的评分：
-
10%
-
20%
-
30%
-
40%
-
50%