工业数据驱动的加氢裂化装置多工况切换过渡状态检测

doi:10.11949/0438-1157.20230729

摘要/Abstract

摘要：

加氢裂化装置运行工况众多且切换频繁，而不同稳定的工况间切换必然存在过渡状态，可能会引起装置运行状态波动，甚至引发事故。通常，操作员会基于专家经验判断装置当前处于稳定或过渡状态，分别采取相应的监控和调整策略。然而，人工判断具有个体差异、经验积累周期长等不足，可能导致过渡状态判断不够准确，故提出一种加氢裂化装置多工况切换过渡状态检测方法，首先，结合工业大数据和装置过程机理，针对工业采集数据运用了小波降噪和平滑，再利用相关分析法和主元分析（principal component analysis，PCA）进行数据降维，剥离了相关性强的变量所带来的额外计算成本和信息干扰，将滑动窗拆分并计算移动方差，再与K-means（K均值）聚类相结合，实现了加氢裂化装置的过渡态检测。最后，与经典K-means聚类和层次聚类方法进行对比验证，证明了所提方法具有更好检测能力。

关键词: 加氢裂化装置, 过渡状态, 主元分析, 移动方差, 集成, 算法, 过程控制

Abstract:

The hydrocracking unit has many operating conditions and frequent switching. There must be a transition state when switching between different stable working conditions, which may cause fluctuations in the operating status of the unit and even cause accidents. Usually, experts will judge that the hydrocracking unit is currently in a stable or transitional state based on expert experience, and adopt corresponding monitoring and adjustment strategies, respectively. However, manual judgment has shortcomings due to individual differences and long experience accumulation period, which may lead to inaccurate transition state judgment. Therefore, a transition state detection method for multi-mode switching of the hydrocracking unit is proposed in this paper. First, combined with industrial big data and device process mechanism, wavelet based noise reduction and smoothing are used for industrial data collection, and then correlation analysis and principal component analysis (PCA) are used to reduce data dimensionality, which extracts the extra highly correlated variables that increase calculation cost and decrease information interference. Combining moving window split and moving variance computation with K-means clustering, transition state detection of the hydrocracking unit is realized. Finally, compared with classical K-means clustering and hierarchical clustering, the proposed method has better performance on transition state detection.

Key words: hydrocracking unit, transition state, principal component analysis, moving variance, integration, algorithm, process control

中图分类号:

TP 277

曹跃, 余冲, 李智, 杨明磊. 工业数据驱动的加氢裂化装置多工况切换过渡状态检测[J]. 化工学报, 2023, 74(9): 3841-3854.

Yue CAO, Chong YU, Zhi LI, Minglei YANG. Industrial data driven transition state detection with multi-mode switching of a hydrocracking unit[J]. CIESC Journal, 2023, 74(9): 3841-3854.

图/表 21

图1 加氢裂化装置简图

Fig.1 Brief flowchart of the hydrocracking unit

表1 不同阈值方法去噪后的聚类轮廓系数

Table 1 Cluster silhouette coefficient after denoising by different threshold methods

方法	最佳簇数下的最佳轮廓系数
heursure	0.8562
rigrsure	0.8427
sqtwolog	0.8565
minimaxi	0.8550

图2 分解层数为5、4、3的小波去噪后聚类簇内距离之和对比

Fig.2 Comparison of the sum of distances in the cluster after the decomposition of 5, 4 and 3 layers

图3 原曲线与降噪平滑处理后曲线对比

Fig.3 Comparison between the original curve and the smoothed curve after noise reduction

表2 不同平滑方法处理信号与原信号相关性对比

Tabel 2 Comparison between processed and original signals by different smoothing methods

方法	相关性系数
Gaussian	0.9269
movmean	0.893

图4 斯皮尔曼相关P值矩阵热图

Fig.4 Heat map of Spearman correlation P-value matrix

图5 累积贡献率

Fig.5 Cumulative contribution

图6 加权主成分

Fig.6 Weighted principal component

图7 基于经典滑动窗的移动方差（以第1860个采样点为例）

Fig.7 Classical moving window based moving variance (the 1860th sample as an example)

图8 基于经典滑动窗的移动方差（以第1252个采样点为例）

Fig.8 Classical moving window based moving variance (the 1252nd sample as an example)

图9 基于滑动窗拆分的移动方差（以第1860个采样点为例）

Fig.9 Moving window split based moving variance (the 1860th sample as an example)

图10 基于滑动窗拆分的移动方差（以第1252个采样点为例）

Fig.10 Moving window split based moving variance (the 1252nd sample as an example)

图11 滑动窗口大小为50的过渡状态检测结果

Fig.11 Results of transition state detection with sliding window size of 50

图12 滑动窗口为80的过渡状态检测结果

Fig.12 Results of transition state detection with sliding window size of 80

图13 滑动窗口为120的过渡状态检测结果

Fig.13 Results of transition state detection with sliding window size of 120

图14 滑动窗口大小为50的聚类效果

Fig.14 Clustering performances with sliding window size of 50

图15 滑动窗口大小为80的聚类效果

Fig.15 Clustering performances of sliding window size of 80

图16 滑动窗口大小为120的聚类效果

Fig.16 Clustering performances with sliding window size of 120

图17 1~4000采样时间聚类结果对比

Fig.17 Comparison of clustering results with the 1—4000 samples

图18 4000~9985采样时间聚类结果对比

Fig.18 Comparison of clustering results with the 4000—9985 samples

表3 过渡状态检测结果

Table 3 Detection results of transition states

方法	过渡状态1	过渡状态2	过渡状态3	过渡状态4
经典K-means聚类	901~1140	3375~3383	3355~3366	6578~6747
	1189~1226	3431~3441	3452~3465	6938~6956
	1889~2013	4060~4084	3491~3562	7217~7193
	2641~3354	4426~4439	3846~3930	7172~7268
	3466~3490	4939~4967	3970~4041	7280~7422
	3563~3845	5360~5379	4451~4500	7513~7533
	9306~9798	5462~5480	4689~4741	7624~7730
		6171~6370	4986~5009	7987~8085
		6485~6497	5205~5334	8461~8486
		6869~6924	5548~5559	8532~8655
		7544~7582	5767~5812	8708~8788
		8113~8157	5991~6015	8885~8918
		9227~9255	6090~6142
			9282~9305
层次聚类	900~1142	3376~3382	3355~3375	6398~6405
	1190~1228	3432~3440	3441~3466	6534~6546
	1887~2015	4063~4082	3489~3563	6571~6751
	2639~3354	4427~4437	3845~4062	6937~6953
	3467~3488	4943~4964	4438~4942	7175~7195
	3564~3844	5362~5377	4970~5361	7214~7415
	9307~9799	5464~5478	5479~6174	7517~7534
		6175~6367	9253~9306	7619~7739
		6871~6922		7985~8087
		7546~7579		8454~8490
		8116~8152		8527~8658
		9229~9252		8703~8799
				8881~8920
方法本文所提	850~1256	3310~3591	3842~4097	6821~6982
	1845~2040	9166~9344	4382~4518	7140~7201
	2613~2673		5333~5506	7486~8189
	9762~9902		6151~6529

参考文献 36

1	姚立松, 许红香. 加氢裂化装置在压减柴油中的作用分析[J]. 石油炼制与化工, 2023, 54(3): 20-24.
	Yao L S, Xu H X. Function analysis of hydrocracking unit in reducing diesel production[J]. Petroleum Processing and Petrochemicals, 2023, 54(3): 20-24.
2	徐春明, 杨朝合. 石油炼制工程[M]. 4版. 北京: 石油工业出版社, 2009.
	Xu C M, Yang Z H. Petroleum Refining Engineering[M]. 4th ed. Beijing: Petroleum Industry Press, 2009.
3	Cao Y, Jan N M, Huang B, et al. Multimodal process monitoring based on variational Bayesian PCA and Kullback-Leibler divergence between mixture models[J]. Chemometrics and Intelligent Laboratory Systems, 2021, 210: 104230.
4	王啸. 多模态工业过程的过渡模态故障监测方法研究[D]. 沈阳: 东北大学, 2015.
	Wang X. Research on monitoring methods about transition process in multi-modal industrial process[D]. Shenyang: Northeastern University, 2015.
5	Zhang K, Peng K X, Zhao S S, et al. A novel feature-extraction-based process monitoring method for multimode processes with common features and its applications to a rolling process[J]. IEEE Transactions on Industrial Informatics, 2021, 17(9): 6466-6475.
6	Jiang Q C, Yan X F. Multimode process monitoring using variational Bayesian inference and canonical correlation analysis[J]. IEEE Transactions on Automation Science and Engineering, 2019, 16(4): 1814-1824.
7	葛志强. 复杂工况过程统计监测方法研究[D]. 杭州: 浙江大学, 2009.
	Ge Z Q. Statistical process monitoring methods for complex processes[D]. Hangzhou: Zhejiang University, 2009.
8	Huang K K, Wei K, Li F B, et al. LSTM-MPC: a deep learning based predictive control method for multimode process control[J]. IEEE Transactions on Industrial Electronics, 2023, 70(11): 11544-11554.
9	Watil A, El Magri A, Lajouad R, et al. Multi-mode control strategy for a stand-alone wind energy conversion system with battery energy storage[J]. Journal of Energy Storage, 2022, 51: 104481.
10	周东华, 王庆林. 基于模型的控制系统故障诊断技术的最新进展[J]. 自动化学报, 1995, 21(2): 244-248.
	Zhou D H, Wang Q L. The latest development of model based fault diagnostics technique of control systems[J]. Acta Automatica Sinica, 1995, 21(2): 244-248.
11	Li W, Zhong X, Shao H D, et al. Multi-mode data augmentation and fault diagnosis of rotating machinery using modified ACGAN designed with new framework[J]. Advanced Engineering Informatics, 2022, 52: 101552.
12	Kodamana H, Raveendran R, Huang B. Mixtures of probabilistic PCA with common structure latent bases for process monitoring[J]. IEEE Transactions on Control Systems Technology, 2017, 27(2): 838-846.
13	Yu J, Qin S J. Multimode process monitoring with Bayesian inference-based finite Gaussian mixture models[J]. AIChE Journal, 2008, 54(7): 1811-1829.
14	Song B, Shi H B, Ma Y X, et al. Multisubspace principal component analysis with local outlier factor for multimode process monitoring[J]. Industrial & Engineering Chemistry Research, 2014, 53(42): 16453-16464.
15	周新杰, 王建林, 艾兴聪, 等. 基于IDPC-RVM的多模态间歇过程质量变量在线预测[J]. 化工学报, 2022, 73(7): 3120-3130.
	Zhou X J, Wang J L, Ai X C, et al. IDPC-RVM based online prediction of quality variables for multimode batch processes[J]. CIESC Journal, 2022, 73(7): 3120-3130.
16	刘伟旻, 王建林, 邱科鹏, 等. 基于DHSC的多模态间歇过程测量数据异常检测方法[J]. 化工学报, 2017, 68(11): 4201-4207.
	Liu W M, Wang J L, Qiu K P, et al. Method for detecting abnormal data in multimode batch processes based on dynamic hypersphere structure change[J]. CIESC Journal, 2017, 68(11): 4201-4207.
17	顾幸生, 周冰倩. 基于LNS-DEWKECA算法的多模态工业过程故障检测[J]. 控制与决策, 2020, 35(8): 1879-1886.
	Gu X S, Zhou B Q. Multimodal industrial process fault detection based on LNS-DEWKECA[J]. Control and Decision, 2020, 35(8): 1879-1886.
18	Cao Y, Jan N M, Huang B, et al. No-delay multimodal process monitoring using Kullback-Leibler divergence-based statistics in probabilistic mixture models[J]. IEEE Transactions on Automation Science and Engineering, 2023, 20(1): 167-178.
19	徐莹, 邓晓刚, 钟娜. 基于ICA混合模型的多工况过程故障诊断方法[J]. 化工学报, 2016, 67(9): 3793-3803.
	Xu Y, Deng X G, Zhong N. A fault diagnosis method for multimode processes based on ICA mixture models[J]. CIESC Journal, 2016, 67(9): 3793-3803.
20	朱红林, 王帆, 侍洪波, 等. 基于非负矩阵分解的多模态过程故障监测方法[J]. 化工学报, 2016, 67(5): 1973-1981.
	Zhu H L, Wang F, Shi H B, et al. Fault detection method based on non-negative matrix factorization for multimode processes[J]. CIESC Journal, 2016, 67(5): 1973-1981.
21	Fang M Q, Kodamana H, Huang B, et al. A novel approach to process operating mode diagnosis using conditional random fields in the presence of missing data[J]. Computers & Chemical Engineering, 2018, 111: 149-163.
22	李元, 杨东昇, 赵丽颖, 等. 层次变分高斯混合模型与主多项式分析的故障检测策略[J]. 化工学报, 2021, 72(3): 1616-1626.
	Li Y, Yang D S, Zhao L Y, et al. Fault detection using hierarchical variational Gaussian mixture model and principal polynomial analysis[J]. CIESC Journal, 2021, 72(3): 1616-1626.
23	张淑美, 王福利, 谭帅, 等. 多模态过程的全自动离线模态识别方法[J]. 自动化学报, 2016, 42(1): 60-80.
	Zhang S M, Wang F L, Tan S, et al. A fully automatic offline mode identification method for multi-mode processes[J]. Acta Automatica Sinica, 2016, 42(1): 60-80.
24	Zheng Y, Wang Y, Yan H L, et al. Density peaks clustering-based steady/transition mode identification and monitoring of multimode processes[J]. The Canadian Journal of Chemical Engineering, 2020, 98(10): 2137-2149.
25	Bakar Z A, Mohemad R, Ahmad A, et al. A comparative study for outlier detection techniques in data mining[C]//2006 IEEE Conference on Cybernetics and Intelligent Systems. Bangkok, Thailand: IEEE, 2006: 1-6.
26	张淑清, 王力, 李昕. 小波分析在故障监测及诊断中的应用[J]. 传感技术学报, 2001, 14(1): 49-53.
	Zhang S Q, Wang L, Li X. The application of wavelet analysis in fault monitor and diagnosis[J]. Journal of Transcluction Technology, 2001, 14(1): 49-53.
27	孔国杰, 张培林, 曹建军, 等. 基于提升小波变换的信号降噪及其工程应用[J]. 计算机工程与应用, 2008, 44(10): 234-237.
	Kong G J, Zhang P L, Cao J J, et al. Signal denoising based on lifting wavelet transform and its application[J]. Computer Engineering and Applications, 2008, 44(10): 234-237.
28	Ballabio D. A MATLAB toolbox for principal component analysis and unsupervised exploration of data structure[J]. Chemometrics and Intelligent Laboratory Systems, 2015, 149: 1-9.
29	王千, 王成, 冯振元, 等. K-means聚类算法研究综述[J]. 电子设计工程, 2012, 20(7): 21-24.
	Wang Q, Wang C, Feng Z Y, et al. Review of K-means clustering algorithm[J]. Electronic Design Engineering, 2012, 20(7): 21-24.
30	管彦周, 万生鹏, 程亚楠, 等. 基于移动方差平均算法的相位敏感光时域反射计去噪算法研究[J]. 仪器仪表学报, 2022, 43(10): 233-240.
	Guan Y Z, Wan S P, Cheng Y N, et al. Research on the denoising algorithm of phase sensitive optical time domain reflectometry based on the moving variance average algorithm[J]. Chinese Journal of Scientific Instrument, 2022, 43(10): 233-240.
31	姜波. 基于MATLAB的小波降噪研究[J]. 电子制作, 2019, 13: 87-88, 99.
	Jiang B. Research on MATLAB based wavelet denoising[J]. Practical Electronics, 2019, 13: 87-88, 99.
32	李加升, 黄文清, 戴瑜兴. 基于自定义阈值函数的小波去噪算法[J]. 电力系统保护与控制, 2008, 36(19): 21-24.
	Li J S, Huang W Q, Dai Y X. Wavelet-based power quality disturbances de-noising by customized thresholding[J]. Power System Protection and Control, 2008, 36(19): 21-24.
33	Donoho D L. Progress in Wavelet Analysis and WVD: A 10-Minute Tour[M]//Progress in Wavelet Analysis and Applications. Gif-sur-Yvette: Editions Frontières. 1993.
34	李蝉娟. 高维数据降维处理关键技术研究[D]. 成都: 电子科技大学, 2017.
	Li C J. Research on key technologies of dimensionality reduction of high-dimensional data[D]. Chengdu: University of Electronic Science and Technology of China, 2017.
35	孙吉贵, 刘杰, 赵连宇. 聚类算法研究[J]. 软件学报, 2008, 19(1): 48-61.
	Sun J G, Liu J, Zhao L Y. Clustering algorithms research[J]. Journal of Software, 2008, 19(1): 48-61.
36	李先鹏, 吴若男, 王义洋, 等. 融合滑动窗口和MLP-AdaBoost的电力负荷预测[J]. 计算机与数字工程, 2023, 51(1): 66-73.
	Li X P, Wu R N, Wang Y Y, et al. Electric arc furnace load forecasting based on sliding window and MLP-AdaBoost[J]. Computer & Digital Engineering, 2023, 51(1): 66-73.

[1]	陈哲文, 魏俊杰, 张玉明. 超临界水煤气化耦合SOFC发电系统集成及其能量转化机制[J]. 化工学报, 2023, 74(9): 3888-3902.
[2]	康飞, 吕伟光, 巨锋, 孙峙. 废锂离子电池放电路径与评价研究[J]. 化工学报, 2023, 74(9): 3903-3911.
[3]	诸程瑛, 王振雷. 基于改进深度强化学习的乙烯裂解炉操作优化[J]. 化工学报, 2023, 74(8): 3429-3437.
[4]	尹刚, 李伊惠, 何飞, 曹文琦, 王民, 颜非亚, 向禹, 卢剑, 罗斌, 卢润廷. 基于KPCA和SVM的铝电解槽漏槽事故预警方法[J]. 化工学报, 2023, 74(8): 3419-3428.
[5]	余娅洁, 李静茹, 周树锋, 李清彪, 詹国武. 基于天然生物模板构建纳米材料及集成催化剂研究进展[J]. 化工学报, 2023, 74(7): 2735-2752.
[6]	李贵贤, 曹阿波, 孟文亮, 王东亮, 杨勇, 周怀荣. 耦合固体氧化物电解槽的CO₂制甲醇过程设计与评价研究[J]. 化工学报, 2023, 74(7): 2999-3009.
[7]	王光, 单发顺, 钱禹丞, 焦建芳. 基于集成学习传递熵的化工过程微小故障检测方法[J]. 化工学报, 2023, 74(7): 2967-2978.
[8]	周继鹏, 何文军, 李涛. 异形催化剂上乙烯催化氧化失活动力学反应工程计算[J]. 化工学报, 2023, 74(6): 2416-2426.
[9]	高学金, 姚玉卓, 韩华云, 齐咏生. 基于注意力动态卷积自编码器的发酵过程故障监测[J]. 化工学报, 2023, 74(6): 2503-2521.
[10]	苏晓丹, 朱干宇, 李会泉, 郑光明, 孟子衡, 李防, 杨云瑞, 习本军, 崔玉. 湿法磷酸半水工艺考察与石膏结晶过程研究[J]. 化工学报, 2023, 74(4): 1805-1817.
[11]	郑书闽, 郭鹏程, 颜建国, 王帅, 李文博, 周淇. 微小通道内过冷流动沸腾阻力特性实验及预测研究[J]. 化工学报, 2023, 74(4): 1549-1560.
[12]	张生安, 刘桂莲. 高效太阳能电解水制氢系统及其性能的多目标优化[J]. 化工学报, 2023, 74(3): 1260-1274.
[13]	张中秋, 李宏光, 石逸林. 基于人工预测调控策略的复杂化工过程多任务学习方法[J]. 化工学报, 2023, 74(3): 1195-1204.
[14]	张江淮, 赵众. 碳三加氢装置鲁棒最小协方差约束控制及应用[J]. 化工学报, 2023, 74(3): 1216-1227.
[15]	顾学荣, 刘硕士, 杨思宇. 基于并行EGO和代理模型辅助的多参数优化方法研究[J]. 化工学报, 2023, 74(3): 1205-1215.