为了提高转炉炼钢自动控制的水平，将理论法、增量计算法和神经元法相结合开发出了炼钢转炉综合智能型静态控制模型。在实际运行中对转炉神经网络模型进行结构优化，对增量模型参数进行实际数据修正，增加参考炉炉数，采用理论与经验模型防止神经网络模型预报结果偏离等改进措施，提高了模型的可靠性和实用性。该模型在转炉自动化炼钢实际生产中获得了良好应用，满足了生产需要。
In order to enhance the level of automatic control of steelmaking in concerter, comprehensive and intelligent models for static control in steelmaking converter have been developed with theory, increment, and neural network methods. The models have been improved during application in terms of structure optimizing of neural network, parameter modification of increment model, increasing heats of reference, and avoiding diverge of neural network forecasting with the theoretical and experiential models. As a result, the reliability and practicality of the models are improved and successfully applied in production of automatic steelmaking.
参考文献
| [1] | 杨立红,刘浏,何平.基于自适应模糊神经网络系统的转炉终点磷的预报控制模型[J].钢铁研究学报,2002(04):47-51. |
| [2] | T.Makino;S.Omiya.Automatic operation of converters at mizushima works[A].,1994:67-71. |
| [3] | 李雨膏.国外氧气转炉计算机控制的新发展上[J].冶金自动化,1987(03):29-34. |
| [4] | 李雨膏.国外氧气转炉计算机控制的新发展下[J].冶金自动化,1987(04):23-28. |
| [5] | 童朝南;张宏伟.冶金生产过程计算机控制[M].北京:冶金工业出版社,1993 |
| [6] | 陈国良.人工神经网络理论进展[J].电子学报,1996(01):70一75. |
| [7] | 程湘君.神经网络原理及其应用[M].北京:国防工业出版社,1995 |
| [8] | 阎平凡.人工神经网络一模型分析与应用[M].合肥:安徽教育出版社,1993 |
| [9] | 王铁.BP算法中学习率及形状因子对学习速度的综合影响[J].上海交通大学学报,1997(03):109-112. |
- 下载量()
- 访问量()
- 您的评分：
-
10%
-
20%
-
30%
-
40%
-
50%