Modeling and application of ethylene cracking furnace based on cross-iterative BLSTM network

doi:10.11949/j.issn.0438-1157.20181373

Abstract

Abstract:

The ethylene cracking furnace model based on BP and RBF ignores the reaction mechanism of cracking furnace and has the disadvantage of large prediction error. Therefore, a two-way long-term time-memory network (BLSTM) model based on the reaction mechanism of ethylene cracking furnace to forecast key parameters such as ethylene yield is proposed. To solve the problem that BLSTM modeling lacks available data, a BLSTM model using cross iteration (CIBLSTM) is provided. The CIBLSTM model uses a forward-reverse cross-iteration method to gradually approximate the true value of the lacks available data, and then a prediction model is established for the ethylene cracking furnace. To verify the validity of the proposed CIBLSTM model, nine industrial actual raw materials and analytical data were selected for simulation test. The simulation results verify the validity and practicability of the proposed CIBLSTM model. The proposed method can also be applied to other complex chemical processes modeling.

Key words: deep learning, neural network, cracking furnace, model, training, prediction

摘要：

数据驱动的乙烯裂解炉模型通常忽视了裂解炉结构和反应机理，存在预测误差偏大的缺点，为此提出基于知识和数据融合驱动的乙烯裂解炉乙烯收率等关键参数的双向长短时间记忆网络(BLSTM)预测模型。为了解决BLSTM建模缺少可用数据的问题，提出了一种采用交叉迭代的BLSTM(CIBLSTM)模型。所提CIBLSTM模型采用了正反向交叉迭代方法，逐步逼近所缺数据的真实值，进而建立乙烯裂解炉的预测模型。为了验证所提CIBLSTM模型的有效性，选取9种工业实际原料与分析数据进行仿真测试，仿真结果验证了所提的CIBLSTM模型的有效性与实用性，所提方法也可应用于其他复杂化工过程建模。

关键词: 深度学习, 神经网络, 裂解炉, 模型, 训练, 预测

CLC Number:

TQ 021.8

Hengchang GU, Peng MU, Jianwei LI. Modeling and application of ethylene cracking furnace based on cross-iterative BLSTM network[J]. CIESC Journal, 2019, 70(2): 548-555.

顾恒昌, 牟鹏, 李建伟. 基于交叉迭代BLSTM网络的乙烯裂解炉建模[J]. 化工学报, 2019, 70(2): 548-555.

Figures/Tables 7

References 30

1	Kumar P , Kunzru D . Modeling of naphtha pyrolysis[J]. Industrial & Engineering Chemistry Process Design & Development, 1985, 24: 774-782.
2	Sadrameli S M . Thermal/catalytic cracking of hydrocarbons for the production of olefins: a state-of-the-art review (I): Thermal cracking review[J]. Fuel, 2015,140: 102-115.
3	Keyvanloo K , Sedighi M , Towfighi J . Genetic algorithm model development for prediction of main products in thermal cracking of naphtha: comparison with kinetic modeling[J]. Chemical Engineering Journal, 2012, 209: 255-262.
4	Hua F , Zhou F , Tong Q . Modeling ethylene cracking process by learning convolutional neural networks[J]. Computer Aided Chemical Engineering, 2018, 44: 841-846.
5	Yuan B , Li J , Du W , et al . Study on co-cracking performance of different hydrocarbon mixture in a steam pyrolysis furnace[J]. Chinese Journal of Chemical Engineering, 2016, 24: 1252-1262.
6	Zhou F , Qiu T , Chen B Z . Analyzing and modeling ethylene cracking process with complex networks approach[J]. Computer Aided Chemical Engineering, 2015, 37: 407-412.
7	Zhou F , Qiu T , Chen B Z . Improvement of ethylene cracking reaction network with network flow analysis algorithm[J]. Computers & Chemical Engineering, 2016, 91: 182-194.
8	Hu G H .Coupled simulation of convection section with dual stage steam feed mixing of an industrial ethylene cracking furnace[J]. Chemical Engineering Journal, 2016, 286:436-446.
9	高荣华 .《化工原理》课程中的微分方程问题[J]. 扬州职业大学学报, 1999, 3(3):45-47.
	Gao R H . Differential equation in the course of chemical engineering theory[J]. Journal of Yangzhou Polytechnic College, 1999, 3(3): 45-47.
10	沙利, 张红梅, 高金森，等 . 乙烯裂解炉管内流动反应历程的数值模拟(Ⅰ)：二维流动反应数学模型的建立[J]. 化工学报, 2003, 54(3): 392-397.
	Sha L , Zhang H M , Gao J S , et al . Numerical simulation on flow-reaction progress in tubular reactor of steam crackers(Ⅰ):Develoment of two-dimensional flow-reaction model[J]. Journal of Chemical Industry and Engineering(China), 2003, 54(3): 392-397.
11	华丰, 方舟, 邱彤 . 乙烯裂解炉反应与传热耦合的智能混合建模与模拟[J]. 化工学报, 2018, 69(3): 923-930.
	Hua F , Fang Z , Qiu T . Recirculation and reaction hybrid intelligent modeling and simulation for industrial ethylene cracking furnace[J]. CIESC Journal, 2018, 69(3): 923-930.
12	Williams G , Baxter R , He H X , et al . A comparative study of RNN for outlier detection in data mining[C]//IEEE International Conference on Data Mining, 2002. Proceedings, Maebashi, Japan: IEEE, 2002: 709-712.
13	Sak H , Senior A , Beaufays F . Long short-term memory recurrent neural network architectures for large scale acoustic modeling[C]//Fifteenth Annual Conference of the International Speech Communication Association. Singapore: Interspeech, 2014.
14	Greff K , Srivastava R K , Koutnik J , et al . LSTM: a search space odyssey[J]. IEEE Transactions on Neural Networks & Learning Systems, 2015, 28: 2222-2232.
15	Sainath T N , Vinyals O , Senior A , et al . Convolutional, long short-term memory, fully connected deep neural networks[C]//Acoustics, Speech and Signal Processing (ICASSP), 2015 IEEE International Conference on. IEEE, Brisbane, QLD, Australia: IEEE, 2015: 4580-4584.
16	彭敏, 杨绍雄, 朱佳晖 . 基于双向LSTM语义强化的主题建模[J]. 中文信息学报, 2018, 32(4): 40-49.
	Peng M , Yang S X , Zhu J H . Semantic enhanced topic modeling by bi-directional LSTM[J]. Journal of Chinese Information Progressing, 2018, 32(4): 40-49.
17	李泽龙, 杨春节, 刘文辉, 等 . 基于LSTM-RNN模型的铁水硅含量预测[J]. 化工学报, 2018，69(3): 992-997.
	Li Z L , Yang C J , Liu W H , et al . Research on hot metal Si-content prediction based on LSTM-RNN[J]. CIESC Journal, 2018, 69(3): 992-997.
18	Zhang X , Zou Y , Li S , et al . Product yields forecasting for fccu via deep bi-directional LSTM network[C] //2018 37th Chinese Control Conference (CCC). IEEE, 2018: 8013-8018.
19	Lefebvre G , Berlemont S , Mamalet F , et al . BLSTM-RNN based 3D gesture classification[C]//International Conference on Artificial Neural Networks. Heidelberg, Berlin: Springer, 2013: 381-388.
20	Hakkani-Tür D , Tür G , Celikyilmaz A , et al . Multi-domain joint semantic frame parsing using bi-directional RNN-LSTM[C]//Proceedings of the 17th Annual Meeting of the International Speech Communication Association (INTERSPEECH 2016). Baixas, France: ISCA 2016: 715-719.
21	Yao Y , Huang Z . Bi-directional LSTM recurrent neural network for Chinese word segmentation[C]//International Conference on Neural Information Processing. Cham: Springer, 2016: 345-353.
22	Wang P , Qian Y , Zhao H , et al . Learning distributed word representations for bidirectional LSTM recurrent neural network[C]//Proceedings of the 2016 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. 2016: 527-533.
23	马涛 . 组合预测方法及其应用研究[D]. 兰州: 兰州大学, 2017.
	Ma T . A research on combining forecasting methods and its applications[D]. Lanzhou: Lanzhou University, 2017.
24	韩昭蓉, 黄廷磊, 任文娟, 等 . 基于Bi-LSTM模型的轨迹异常点检测算法[J]. 雷达学报, 2018, (5):1-8.
	Han Z R , Huang T L , Ren W J ，et al . Trajectory outlier detection algorithm based on bi-LSTM Model[J].Journal of Radars, 2018, (5): 1-8.
25	张利军, 张永刚, 王国清, 等 . 石脑油的组成预测方法[J]. 化工进展, 2011, 30(2): 278-283.
	Zhang L J , Zhang Y G , Wang G Q , et al . Rearch on prediciton of naphtha composition based on commerical indices[J]. Chemical Industry and Engineering Progress, 2011, 30(2): 278-283.
26	张杰, 黄翔, 李泷杲，等 . 基于T-scan测量的薄壁钣金件孔特征重构[J]. 工程科学学报, 2017, 39(6): 917-923.
	Zhang J , Huang X , Li S G . Hole feature reconstruction of thin-walled sheet metal parts based on T-scan measurement[J]. Chinese Journal of Engineering, 2017, 39(6): 917-923.
27	王飞 . 乙烯装置中碳二加氢反应器的稳态、动态建模与优化[D]. 上海: 华东理工大学, 2012.
	Wang F . Steady-state, dynamic simulation and optimization of acetylene hydrogenation reactor in ethylene plant[D]. Shanghai: East China University of Science and Technolog, 2012.
28	Prokhorov D V , Si J , Barto A , et al . Handbook of Learning and Approximate Dynamic Programming: Chapter 15 BPTT and DAC -A Common Framework for Comparison[M]. USA: John Wiley & Sons, Inc., Online library,2012.
29	邵研硕 . 乙烯生产过程能效监测系统预测评估模块实现[D]. 大连: 大连理工大学, 2017.
	Zhao Y S . The implementation of forecast evaluation module in the ethylene production process inspection system of energy efficiency[D]. Dalian: Dalian University of Technology, 2017.
30	耿志强, 陈杰, 韩永明 . 基于模糊RBF神经网络的乙烯装置生产能力预测[J]. 化工学报, 2016, 67(3): 812-819.
	Geng Z Q , Chen J , Han Y M . Ethylenen plants production capacity forecast based on fuzzy RBF neural network[J]. CIESC Journal, 2016, 67(3): 812-819.

Structure parameters	Value	Operation parameters	Value
furnace tube group	6	feedstock flow/(kg/h)	890.625
tube pass	2	steam/hydrocarbon ratio	0.60
arrangement	16/8	coil inlet temperature/K	875
inner diameter/m	0.051/0.073	coil outlet temperature/K	1122
outer diameter/m	0.063/0.086	coil outlet pressure/kPa	178
length/m	13.681/14.921
tube pitch/m	0.112/0.154

Structure parameters	Value	Operation parameters	Value
furnace tube group	6	feedstock flow/(kg/h)	890.625
tube pass	2	steam/hydrocarbon ratio	0.60
arrangement	16/8	coil inlet temperature/K	875
inner diameter/m	0.051/0.073	coil outlet temperature/K	1122
outer diameter/m	0.063/0.086	coil outlet pressure/kPa	178
length/m	13.681/14.921
tube pitch/m	0.112/0.154

No.	Density，20℃/(g/cm³)	10% boiling range/℃	30% boiling range/℃	50% boiling range/℃	70% boiling range/℃	90% boiling range/℃	P	N	A	L_mn
1	0.6927	54.5	73	90	107	127	0.7223	0.2234	0.0529	0
2	0.7255	86.9	105.4	121.3	137.3	156.5	0.663	0.2605	0.0762	1.382
3	0.7488	85.2	114.8	131.8	147.5	162.4	0.4631	0.4856	0.0483	1.7751
4	0.7022	60	75	88.5	103	124	0.7069	0.2248	0.0566	2.1174
5	0.7103	64	81	98	117	143	0.6669	0.2605	0.068	2.5121
6	0.7389	77	106	135	163	190	0.6898	0.1916	0.116	2.9868
7	0.7093	69	88	103	119	149	0.6409	0.3172	0.0419	0.0023
8	0.7375	85.2	105	120.3	136.2	156.1	0.5609	0.3493	0.0893	0.0092
9	0.6894	36.2	46.9	65	104.7	137.6	0.7379	0.1742	0.0864	0.0133

No.	Density，20℃/(g/cm³)	10% boiling range/℃	30% boiling range/℃	50% boiling range/℃	70% boiling range/℃	90% boiling range/℃	P	N	A	L_mn
1	0.6927	54.5	73	90	107	127	0.7223	0.2234	0.0529	0
2	0.7255	86.9	105.4	121.3	137.3	156.5	0.663	0.2605	0.0762	1.382
3	0.7488	85.2	114.8	131.8	147.5	162.4	0.4631	0.4856	0.0483	1.7751
4	0.7022	60	75	88.5	103	124	0.7069	0.2248	0.0566	2.1174
5	0.7103	64	81	98	117	143	0.6669	0.2605	0.068	2.5121
6	0.7389	77	106	135	163	190	0.6898	0.1916	0.116	2.9868
7	0.7093	69	88	103	119	149	0.6409	0.3172	0.0419	0.0023
8	0.7375	85.2	105	120.3	136.2	156.1	0.5609	0.3493	0.0893	0.0092
9	0.6894	36.2	46.9	65	104.7	137.6	0.7379	0.1742	0.0864	0.0133

物质	目标产量/%(质量)	预测收率/%(质量)	误差/%
H₂	12.36	12.10	2.19
CH₄	22.48	22.27	0.93
C₂H₄	22.54	22.35	0.81
C₂H₆	4.67	4.60	1.38
C₃H₆	9.86	9.86	0.00
C₃H₈	4.42	4.00	10.51
C₄H₁₀	5.32	5.14	3.58
C₄H₈	6.45	6.14	4.94
C₄H₆	3.03	2.82	7.24
C₄'	5.25	5.13	2.36