在传统BP神经网络预测模型的基础上,依据灰色理论中的灰色关联度,提出了输出变量各个影响因素的灰色关联度权值,首次建立基于灰色理论的神经网络预测模型,并依据国内某钢厂300组实际生产数据进行仿真试验。试验结果表明：误差绝对值小于5%的炉数有39炉,占总炉数的65.00%;误差绝对值小于10%的炉数共有58炉,占到96.67%。与传统BP神经网络相比,基于灰色理论的神经网络模型的预测精度提高近12.5%,说明基于灰色理论的铁水预处理终点磷含量神经网络预测模型能更精确地反映现场实际水平。
On the basis of optimizing the traditional BP neural network model,a neural network prediction model combined with grey theory was developed by laying correlative degree weights to all input factors which had effects on the output variable.And then simulation experiments of model newly established were conducted based on data from a domestic steel plant.The results show that hit rate arrives at 65.00% when the error modulus is less than 5.00%,and the value is 96.67% when less than 10.00%.Comparing to the traditional neural network prediction model,the accuracy almost increases by 12.50%,Thus,the prediction of end point phosphorus content fits the real perfectly,which accounts for that neural network model for terminative phosphorus content based on grey theory can reflect accurately the practice in hot metal pretreatment.
参考文献
| [1] | 殷瑞钰.冶金工程流程学[M].北京:冶金工业出版社,2007 |
| [2] | 刘锟,刘浏,何平,刘伟.基于烟气分析转炉终点碳含量控制的新算法[J].炼钢,2009(01):33-37. |
| [3] | 周世纪,彭燕妮,曹长修.基于广义遗传优化的RBF算法在铁水脱硫中的应用[J].重庆大学学报（自然科学版）,2005(02):77-80. |
| [4] | Huixin TIAN;Zhizhong MAO;Yan WANG .Hybrid Modeling of Molten Steel Temperature Prediction in LF[J].ISIJ International,2008(1):58-62. |
| [5] | 田慧欣,毛志忠,王嘉铮.基于ELM新方法的LF终点温度软测量混合模型[J].东北大学学报（自然科学版）,2008(01):33-36. |
| [6] | 王心哲,韩敏.基于变量选择的转炉炼钢终点预报模型[J].控制与决策,2010(10):1589-1592. |
| [7] | 杨为民,赵杰,韦生勇,贾哲.基于混合递阶遗传神经网络的转炉温度预报[J].控制工程,2007(z1):33-34,63. |
| [8] | 刘思峰;郭天榜;党耀国.灰色系统理论及其应用[M].北京:科学出版社,2010 |
| [9] | 丁敬国,曾庆亮,胡贤磊,李西安,刘相华.中厚板轧制过程基于灰色关联度的厚度修正方法[J].钢铁研究学报,2009(11):20-23. |
| [10] | 刘锟,刘浏,何平.基于副枪的转炉终点磷锰控制模型与脱磷优化[J].钢铁,2008(07):32-36. |
| [11] | 张慧书,战东平,姜周华.改进BP网络在铁水预脱硫终点硫含量预报中的应用[J].东北大学学报（自然科学版）,2007(08):1140-1142. |
- 下载量()
- 访问量()
- 您的评分：
-
10%
-
20%
-
30%
-
40%
-
50%