基于ESN模型的焦炉火道温度预测研究
|
摘要
为准确建立焦炉火道温度模型,保证焦炭质量,节约能耗,提出了基于回声状态网络(echo state network,ESN)模型的焦炉火道温度的预测方法。针对炼焦过程中强耦合、大滞后的特点,以及实际生产中大量炼焦数据未被合理利用的现状,运用数据驱动与非线性建模相结合的方法,首先,对采集的数据进行数据处理,保证了数据的真实有效性;然后,分别建立了焦炉火道温度系统的BP神经网络预测模型和ESN预测模型;最后,在Matlab环境下进行仿真实验,并对2种预测模型的平均相对误差和命中率进行对比。实验表明:ESN模型与BP神经网络模型相比,平均相对误差减小了0.66%,命中率提高了6.39%,说明在结合数据驱动的前提下,ESN模型更能准确预测火道温度,为下一步火道温度的优化控制奠定基础。
To accurately establish the temperature model of coke oven flue,ensure coke quality and save energy consumption,the temperature prediction method of coke oven flue based on echo state network(ESN)was proposed.According to the characteristics of coke oven heating process such as large lag,strong coupling,and the status of large amount of production data have not been reasonably utilized,the method of data-driven combined with nonlinear modeling was employed for the study.Firstly,the collected data were analyzed to ensure the validity of the data.Then,the BP neural network prediction model and ESN prediction model of coke oven flue temperature system were established,respectively.Finally,simulation experiments in Matlab were carried out,and the average relative error and hits rate of two prediction models were compared.Experiment results show that comparing with BP neural network model,the average relative error of ESN model decreases 0.66%,and the hits rate increases6.39%,indicating the ESN prediction model can predict flue temperature precisely combining data-driven,which lays a foundation for further optimization control.
To accurately establish the temperature model of coke oven flue,ensure coke quality and save energy consumption,the temperature prediction method of coke oven flue based on echo state network(ESN)was proposed.According to the characteristics of coke oven heating process such as large lag,strong coupling,and the status of large amount of production data have not been reasonably utilized,the method of data-driven combined with nonlinear modeling was employed for the study.Firstly,the collected data were analyzed to ensure the validity of the data.Then,the BP neural network prediction model and ESN prediction model of coke oven flue temperature system were established,respectively.Finally,simulation experiments in Matlab were carried out,and the average relative error and hits rate of two prediction models were compared.Experiment results show that comparing with BP neural network model,the average relative error of ESN model decreases 0.66%,and the hits rate increases6.39%,indicating the ESN prediction model can predict flue temperature precisely combining data-driven,which lays a foundation for further optimization control.
引文
[1]郑明东,宁方青.焦炉立火道温度控制研究[J].大连理工大学学报,2001,41(4):442-445.ZHENG Mingdong,NING Fangqing.Research on coke oven heating control system[J].Journal of Dalian University of Technology,2001,41(4):442-445.(in Chinese)
[2]雷琪,吴敏,曹卫华.焦炉加热过程无辨识自适应控制方法[J].控制工程,2008,15(3):279-282.LEI Qi,WU Min,CAO Weihua.Identification-free adaptive control method for the heating process in coke oven[J].Control Engineering of China,2008,15(3):279-282.(in Chinese)
[3]曹卫华,陈泰任,吴敏,等.基于误差预测的焦炉火道温度软测量模型[J].信息与控制,2009,38(2):206-210.CAO Weihua,CHEN Tairen,WU Min,et al.An error-forecasting-based soft-sensing model for coke oven flue temperature[J].Information&Control,2009,38(2):206-210.(in Chinese)
[4]王伟,吴敏,曹卫华,等.基于组合灰色预测模型的焦炉火道温度模糊专家控制[J].控制与决策,2010,25(2):185-190.WANG Wei,WU Min,CAO Weihua,et al.Fuzzy-expert control based on combination grey prediction model for flue temperature in coke oven[J].Control&Decision,2010,25(2):185-190.(in Chinese)
[5]丁明,王磊,毕锐.基于改进BP神经网络的光伏发电系统输出功率短期预测模型[J].电力系统保护与控制,2012,40(11):93-99.DING Ming,WANG Lei,BI Rui.A short-term prediction model to forecast output power of photovoltaic system based on improved BP neural network[J].Power System Protection&Control,2012,40(11):93-99.(in Chinese)
[6]王丽婕,冬雷,廖晓钟,等.基于小波分析的风电场短期发电功率预测[J].中国电机工程学报,2009,29(28):30-33.WANG Lijie,DONG Lei,LIAO Xiaozhong,et al.Short-term power prediction of a wind farm based on wavelet analysis[J].Proceedings of the CSEE,2009,29(28):30-33.(in Chinese)
[7]李勇,胡双启.灰色Elman神经网络在火灾事故预测中的应用研究[J].中国安全科学学报,2009,19(3):28-31.LI Yong,HU Shuangqi.The application study of gray Elman neural networks to fire accident prediction[J].China Safety Science Journal,2009,19(3):28-31.(in Chinese)
[8]梁栋,张凤琴,陈大武,等.一种基于决策树和遗传算法-BP神经网络的组合预测模型[J].中国科技论文,2015,10(2):169-174.LIANG Dong,ZHANG Fengqin,CHEN Dawu,et al.Composite prediction model based on decision tree and GA-BPNN[J].China.Sciencepaper,2015,10(2):169-174.(in Chinese)
[9]刘颖,时飞飞,赵珺,等.基于改进回声状态网络的高炉煤气产耗预测[J].系统仿真学报,2011,23(10):2184-2189.LIU Ying,SHI Feifei,ZHAO Jun,et al.Study on prediction method for generation and consumption of blast furnace gas based on improved echo state network[J].Journal of System Simulation,2011,23(10):2184-2189.(in Chinese)
[10]韩敏,许美玲,王新迎.多元时间序列的子空间回声状态网络预测模型[J].计算机学报,2014,37(11):2268-2275.HAN Min,XU Meiling,WANG Xinying.A multivariate time series prediction model based on subspace echo state network[J].Chinese Journal of Computers,2014,37(11):2268-2275.(in Chinese)
[11]COWPER M R,MULGREW B,UNSWORTH C P.Nonlinear prediction of chaotic signals using a normalized radial basis function network[J].Signal Processing,2002,82(5):775-789.
[12]楼旭伟,楼辉波,朱剑锋.基于遗传算法径向基神经网络的交通流预测[J].中国科技论文,2013,8(11):1141-1144.LOU Xuwei,LOU Huibo,ZHU Jianfeng.Prediction of traffic flow of optimized radial basis function neural network based on genetic algorithm[J].China Sciencepaper,2013,8(11):1141-1144.(in Chinese)
[13]MLLER K R,SMOLA A J,RTSCH G,et al.Predicting time series with support vector machines[J].Lecture Notes in Computer Science,1997,20(2):999-1004.
[14]MUKHERJEE S,OSUNA E,GIROSI F.Nonlinear prediction of chaotic time series using support vector machines[C]//IEEE Workshop on Neural Networks for Signal Processing VII.[S.l.]:IEEE,1997:511-520.
[15]CAO L J,TAY F E H.Support vector machine with adaptive parameters in financial time series forecasting[J].IEEE Transactions on Neural Networks,2003,14(6):1506-1518.
[16]VESANTO J.Using the SOM and local models in timeseries prediction[C]//Proceedings of Workshop on Self-Organizing Maps(WSOM’97),Espoo,Finland,[s.n.],1997:209-214.
[17]FELDKAMP L A,PROKHOROV D V,EAGEN C F,et al.Enhanced multi-stream Kalman filter training for recurrent networks[M]//Nonlinear Modeling.New York,Springer US,1998:29-53.
[18]PRINCIPE J C,KUO J M.Dynamic modelling of chaotic time series with neural networks[J].Advances in Neural Information Processing Systems,1998,7:311-318.
[19]JAEGER H.Tutorial on training recurrent neural networks,covering BPPT,RTRL,EKF and the echo state network approach[R].GMD Report 159,German National Research Center for Information Technology,2002,48:3-45.
[20]JAEGER H,HAAS H.Harnessing nonlinearity:predicting chaotic systems and saving energy in wireless communication[J].Science,2004,304(5667):78-80.
[21]宋彤,李菡.基于小波回声状态网络的混沌时间序列预测[J].物理学报,2012,61(8):080506-1-080506-7.SONG Tong,LI Han.Chaotic time series prediction based on wavelet echo state network[J].Acta Physica Sinica,2012,61(8):080506-1-080506-7.(in Chinese)
[2]雷琪,吴敏,曹卫华.焦炉加热过程无辨识自适应控制方法[J].控制工程,2008,15(3):279-282.LEI Qi,WU Min,CAO Weihua.Identification-free adaptive control method for the heating process in coke oven[J].Control Engineering of China,2008,15(3):279-282.(in Chinese)
[3]曹卫华,陈泰任,吴敏,等.基于误差预测的焦炉火道温度软测量模型[J].信息与控制,2009,38(2):206-210.CAO Weihua,CHEN Tairen,WU Min,et al.An error-forecasting-based soft-sensing model for coke oven flue temperature[J].Information&Control,2009,38(2):206-210.(in Chinese)
[4]王伟,吴敏,曹卫华,等.基于组合灰色预测模型的焦炉火道温度模糊专家控制[J].控制与决策,2010,25(2):185-190.WANG Wei,WU Min,CAO Weihua,et al.Fuzzy-expert control based on combination grey prediction model for flue temperature in coke oven[J].Control&Decision,2010,25(2):185-190.(in Chinese)
[5]丁明,王磊,毕锐.基于改进BP神经网络的光伏发电系统输出功率短期预测模型[J].电力系统保护与控制,2012,40(11):93-99.DING Ming,WANG Lei,BI Rui.A short-term prediction model to forecast output power of photovoltaic system based on improved BP neural network[J].Power System Protection&Control,2012,40(11):93-99.(in Chinese)
[6]王丽婕,冬雷,廖晓钟,等.基于小波分析的风电场短期发电功率预测[J].中国电机工程学报,2009,29(28):30-33.WANG Lijie,DONG Lei,LIAO Xiaozhong,et al.Short-term power prediction of a wind farm based on wavelet analysis[J].Proceedings of the CSEE,2009,29(28):30-33.(in Chinese)
[7]李勇,胡双启.灰色Elman神经网络在火灾事故预测中的应用研究[J].中国安全科学学报,2009,19(3):28-31.LI Yong,HU Shuangqi.The application study of gray Elman neural networks to fire accident prediction[J].China Safety Science Journal,2009,19(3):28-31.(in Chinese)
[8]梁栋,张凤琴,陈大武,等.一种基于决策树和遗传算法-BP神经网络的组合预测模型[J].中国科技论文,2015,10(2):169-174.LIANG Dong,ZHANG Fengqin,CHEN Dawu,et al.Composite prediction model based on decision tree and GA-BPNN[J].China.Sciencepaper,2015,10(2):169-174.(in Chinese)
[9]刘颖,时飞飞,赵珺,等.基于改进回声状态网络的高炉煤气产耗预测[J].系统仿真学报,2011,23(10):2184-2189.LIU Ying,SHI Feifei,ZHAO Jun,et al.Study on prediction method for generation and consumption of blast furnace gas based on improved echo state network[J].Journal of System Simulation,2011,23(10):2184-2189.(in Chinese)
[10]韩敏,许美玲,王新迎.多元时间序列的子空间回声状态网络预测模型[J].计算机学报,2014,37(11):2268-2275.HAN Min,XU Meiling,WANG Xinying.A multivariate time series prediction model based on subspace echo state network[J].Chinese Journal of Computers,2014,37(11):2268-2275.(in Chinese)
[11]COWPER M R,MULGREW B,UNSWORTH C P.Nonlinear prediction of chaotic signals using a normalized radial basis function network[J].Signal Processing,2002,82(5):775-789.
[12]楼旭伟,楼辉波,朱剑锋.基于遗传算法径向基神经网络的交通流预测[J].中国科技论文,2013,8(11):1141-1144.LOU Xuwei,LOU Huibo,ZHU Jianfeng.Prediction of traffic flow of optimized radial basis function neural network based on genetic algorithm[J].China Sciencepaper,2013,8(11):1141-1144.(in Chinese)
[13]MLLER K R,SMOLA A J,RTSCH G,et al.Predicting time series with support vector machines[J].Lecture Notes in Computer Science,1997,20(2):999-1004.
[14]MUKHERJEE S,OSUNA E,GIROSI F.Nonlinear prediction of chaotic time series using support vector machines[C]//IEEE Workshop on Neural Networks for Signal Processing VII.[S.l.]:IEEE,1997:511-520.
[15]CAO L J,TAY F E H.Support vector machine with adaptive parameters in financial time series forecasting[J].IEEE Transactions on Neural Networks,2003,14(6):1506-1518.
[16]VESANTO J.Using the SOM and local models in timeseries prediction[C]//Proceedings of Workshop on Self-Organizing Maps(WSOM’97),Espoo,Finland,[s.n.],1997:209-214.
[17]FELDKAMP L A,PROKHOROV D V,EAGEN C F,et al.Enhanced multi-stream Kalman filter training for recurrent networks[M]//Nonlinear Modeling.New York,Springer US,1998:29-53.
[18]PRINCIPE J C,KUO J M.Dynamic modelling of chaotic time series with neural networks[J].Advances in Neural Information Processing Systems,1998,7:311-318.
[19]JAEGER H.Tutorial on training recurrent neural networks,covering BPPT,RTRL,EKF and the echo state network approach[R].GMD Report 159,German National Research Center for Information Technology,2002,48:3-45.
[20]JAEGER H,HAAS H.Harnessing nonlinearity:predicting chaotic systems and saving energy in wireless communication[J].Science,2004,304(5667):78-80.
[21]宋彤,李菡.基于小波回声状态网络的混沌时间序列预测[J].物理学报,2012,61(8):080506-1-080506-7.SONG Tong,LI Han.Chaotic time series prediction based on wavelet echo state network[J].Acta Physica Sinica,2012,61(8):080506-1-080506-7.(in Chinese)
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |