在双辊铸轧过程中,铸轧力的控制是铸轧过程稳定进行和提高薄带质量的关键.为了控制铸轧力,必须建立铸轧力计算数学模型,本文采用了一种基于贝叶斯方法的前向神经网络训练算法以提高网络的泛化能力,在网络的目标函数中引入了表示网络结构复杂性的惩罚项,融入"奥克姆剪刀"理论,避免了网络训练的过拟合.将上述网络应用于铸轧过程的铸轧力计算,具有很高的计算精度,同时在收敛速度、稳定性和泛化能力方面都优于传统的BP神经网络.
参考文献
| [1] | K.Schwerdtfeger .Benefits, Challenges and Limits in New Routes for Hot Strip Production[J].ISIJ International,1998(8):852-861. |
| [2] | A L Robson;G L Thompson .Direct Casting of thin strip[J].Materials World,1995(5):222-224. |
| [3] | 邸洪双,鲍培玮,苗雨川,王国栋,刘相华.双辊铸轧薄带钢实验研究及工艺稳定性分析[J].东北大学学报（自然科学版）,2000(03):274-277. |
| [4] | Bernhard S;Enning M;Rake H .Automation of a hboratow plant for direct casting of thin steel strips[J].Control Engineering Practice,1994,2(06):961-967. |
| [5] | 流函数法在铸轧变形理论分析中的应用[J].中国有色金属学报,1999(01):115. |
| [6] | 张铃;张钹.人工神经网络理论及应用[M].杭州:浙江科学技术出版社,1997:50-58. |
| [7] | 王国栋;刘相华.金属轧制过程人工智能优化[M].北京:冶金工业出版社,2000:21-137. |
| [8] | Foresee F D;Hagan M T.Gauss-newton approximation to bayesian learning[A].Houston,Texas,1997 |
| [9] | 魏东,张明廉,蒋志坚,孙明.基于贝叶斯方法的神经网络非线性模型辨识[J].计算机工程与应用,2005(11):5-8,11. |
| [10] | R. Orre;A. Lansner;A. Bate;M. Lindquist .Bayesian neural networks with confidence estimations applied to data mining[J].Computational statistics & data analysis,2000(4):473-493. |
| [11] | Mackay D J c .A practical bayesian framework for backpropagation networks[J].Neural Computation,1992,4(03):448-472. |
| [12] | Penny W D;Roberts S I .Bayesian neural networks for classification:how useful is the evidence framework[J].Neural Networks,1999,12:877-892. |
| [13] | Jia Li;Yinlun Huang .Bayesian-based on-line applicability evaluation of neural network models in modeling automotive paint spray operations[J].Computers & Chemical Engineering: An International Journal of Computer Applications in Chemical Engineering,2006(9):1392-1399. |
| [14] | Hagan M.T.;Menhaj M.B. .Training feedforward networks with the Marquardt algorithm[J].IEEE Transactions on Neural Networks,1994(6):989-993. |
| [15] | 刘兴刚,丛德宏,郝丽娜,徐心和,李山青.基于数据库的神经网络轧制力建模[J].系统仿真学报,2005(01):7-10,15. |
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%