摘要
火道温度稳定在一定范围内是焦炉加热过程优化控制的目的,也是保证焦炭质量、提高产量的前提。针对焦炉加热过程大滞后、强耦合等特点,提出一种基于极限学习机(ELM)的预测算法;考虑到实际生产过程中焦炉机侧与焦侧是分离操作控制的,从而建立两侧的预测模型来分别预测机、焦侧火道温度,增加模型的准确性。实验表明:建立机、焦两侧的ELM预测模型较传统BP网络算法更能准确的预测两侧火道温度,可以为焦炉的优化控制提供良好的指导。
Keeping flue temperature stable within a certain range is not only the purpose of optimal control of coke oven heating process, but also the prerequisite of ensuring the quality of coke and increasing production. According to the characteristics of coke oven heating process such as large lag, strong coupling, a prediction algorithm based on the extreme learning machine(ELM) method is proposed. Considering the operational control of coke oven's machine side and coke side is separate in the actual production process, prediction models of both sides are built to predict the flue temperature of machine side and coke side, which increase the accuracy of the model. Experiment shows that establishing the ELM prediction models of machine side and coke side can predict both sides' flue temperature precisely compared to traditional BP network algorithm, and provide a good guidance for optimal control of coke oven.
引文
1 Zhen W H,Liu H C,Zhou K.Present situation and development of coke production in China.Iron and Steel,2004,39(3):67-73
2 Lei Q,Wu M,Cao W H,et al.An intelligent integrated method for soft-Sensing of the flue temperature in coke oven and its application.Journal of East China University of Science and Technology(Natural Science Edition),2006,32(7):762-766
3 Gao C H,Jian L,Chen J M,et al.Data-driven modeling and predictive algorithm for complex blast furnace ironmaking process.Acta Automatica Sinica,2009,35(6):725-730
4 Zhang Y,Li J,Cui G M.Study of hot iron temperature prediction model using wavelet neural network.Computers and Applied Chemistry,2013,30(10):1173-1176
5 Ding G,Xu M Q,Hou L G.Prediction of aeroengine exhaust gas temperature using process neural network.Journal of Aerospace Power,2009,24(5):1035-1039
6 Cao W H,Du N,An J Q,et al.Prediction method of blast furnace hanging based on fusion of subjective and objective evidences.Journal of University of Science and Technology Beijing,2014,36(4):506-514
7 Li J J,Yang Z Y,Cao Y.Research on hot metal silicon content prediction based on integrated neural network.Computers and Applied Chemistry,2013,30(10):1113-1116
8 Gao G Y,Jiang G P.Prediction of multivariable chaotic time series using optimized extreme learning machine.Acta Phys.Sin,2012,61(4):37-45
9 Huang G B,Zhu Q Y,Siew C K.Extreme Learning Machine:Theory and Application[J].Neurocomputing,2006,70:489-501