基于DBSVDD-RVR的多模态间歇过程质量变量在线软测量

doi:10.11949/0438-1157.20240333

摘要/Abstract

摘要：

现有的多模态间歇过程软测量未考虑过程数据的批次差异及过渡模态的复杂时变特性，影响了间歇过程模态识别的合理性及质量变量在线软测量的准确性。提出了一种基于双边界支持向量数据描述-相关向量回归（double boundary support vector data description-relevance vector regression，DBSVDD-RVR）的间歇过程质量变量在线软测量方法。依据间歇过程离线模态划分获得的各稳定及过渡模态历史数据，建立DBSVDD在线模态识别模型，并引入滑动窗，构建间歇过程在线模态识别策略，利用DBSVDD模型实现在线测量数据的模态识别；在此基础上，构建了基于超球体距离的数据相似度计算方法，选择过渡模态在线数据的相似建模数据集，建立过渡模态的即时学习RVR软测量模型，并依据历史数据建立各稳定模态的RVR软测量模型，实现间歇过程质量变量的在线软测量。青霉素发酵过程的实验结果表明，所提方法有效地提高了间歇过程模态识别的合理性和质量变量在线软测量的准确性。

关键词: 间歇式, 双边界支持向量数据描述, 在线模态识别, 模型, 预测

Abstract:

The existing multimode batch process soft sensor does not consider the batch difference of process data and the complex time-varying characteristics of transition modes, which affects the rationality of batch process mode identification and the accuracy of online soft sensing of quality variables. This paper proposes an online soft sensing method for batch process quality variables based on double boundary support vector data description-relevance vector regression (DBSVDD-RVR). According to the historical data of stable and transition modes obtained by offline mode partition of batch processes, the online mode identification model of DBSVDD was established. Then, the sliding window was introduced to construct the online mode identification strategy, and the online mode identification of batch data was realized by using DBSVDD model. On this basis, the data similarity calculation method based on hypersphere distance was constructed, and the similarity modeling data set of online data in transition mode was selected to establish the just-in-time learning RVR soft sensing model of transition mode. The RVR soft sensing model of each stable mode was established according to the historical data, and the online soft sensing of batch process quality variables was realized. The experimental results of penicillin fermentation process show that the proposed method effectively improves the rationality of mode identification and the accuracy of online soft sensing for quality variables in batch processes.

Key words: batchwise, double boundary support vector data description, online mode identification, model, prediction

中图分类号:

TP 274

李季, 王建林, 何睿, 周新杰, 王雯, 赵利强. 基于DBSVDD-RVR的多模态间歇过程质量变量在线软测量[J]. 化工学报, 2024, 75(9): 3231-3241.

Ji LI, Jianlin WANG, Rui HE, Xinjie ZHOU, Wen WANG, Liqiang ZHAO. DBSVDD-RVR based online soft sensing for quality variables in multimode batch processes[J]. CIESC Journal, 2024, 75(9): 3231-3241.

图/表 10

图1 基于DBSVDD的间歇过程在线模态识别模型构建示意图

Fig.1 Constructing of mode identification model for batch processes based on DBSVDD

图2 基于DBSVDD的间歇过程模态识别方法流程图

Fig.2 The flow chart of mode identification method for batch processes based on DBSVDD

图3 基于DBSVDD-RVR的多模态间歇过程质量变量在线软测量流程图

Fig.3 Flow chart of online soft sensing of quality variables for multimode batch processes based on DBSVDD-RVR

表1 青霉素发酵过程变量

Table 1 Process variables of penicillin fermentation

No.	过程变量（单位）	No.	过程变量（单位）
1	通风率（L/h）	7	二氧化碳浓度（mmol/L）
2	搅拌功率（W）	8	产热量（kcal/h）
3	底物流加速率（L/h）	9	加酸流速（ml/h）
4	底物流温度（K）	10	加碱流速（ml/h）
5	溶解氧浓度（mol/L）	11	青霉素浓度（g/L）
6	反应器体积（L）

表2 各DBSVDD模型的内外边界

Table 2 The inner and outer boundaries of each DBSVDD model

DBSVDD 模型序号	内边界	外边界	构建模型所用模态数据序号
1	0.9283	1.0591	1、2
2	0.9513	1.0334	2、3
3	0.9513	1.0055	3、4
4	0.9840	0.9951	4、5

图4 不同方法对第1个测试批次的在线模态识别结果

Fig.4 Mode identification results of the first online batch

表3 不同方法对5个测试批次的在线模态识别结果

Table 3 Mode identification results for 5 test batches by different methods

方法	测试批次序号	第1模态	第2模态	第3模态	第4模态	第5模态
时间标签	1~5	[1,42]	[43,50]	[51,137]	[138,186]	[187,400]
DBSVDD	1	[1,41]	[42,48]	[49,142]	[143,189]	[190,400]
	2	[1,43]	[44,50]	[51,150]	[151,185]	[186,400]
	3	[1,40]	[41,48]	[49,150]	[151,192]	[193,400]
	4	[1,40]	[41,48]	[49,150]	[151,192]	[193,400]
	5	[1,41]	[42,48]	[49,137]	[138,189]	[190,400]

图5 第1个测试批次不同方法的青霉素浓度在线软测量结果

Fig.5 Online soft sensing results of penicillin concentration for different methods in the first test batch

图6 第1个测试批次不同方法的青霉素浓度在线软测量误差

Fig.6 Online soft sensing error of penicillin concentration for different methods in the first test batch

表4 不同方法对5个测试批次青霉素浓度软测量结果的RMSE和R2

Table 4 RMSE and R2 of penicillin concentration soft sensing results for 5 test batches by different methods

批次序号	DS-RVR		TS-RVR		本文方法
批次序号	RMSE	R²	RMSE	R²	RMSE	R²
1	0.1143	0.9114	0.0194	0.9981	0.0136	0.9991
2	0.1729	0.8715	0.0137	0.9990	0.0078	0.9997
3	0.1451	0.9084	0.0142	0.9990	0.0090	0.9996
4	0.1447	0.9095	0.0239	0.9971	0.0186	0.9982
5	0.1447	0.9056	0.0204	0.9978	0.0152	0.9988
average	0.1518	0.8988	0.0181	0.9982	0.0126	0.9991

参考文献 32

1	Chang P, Lu R W. Process monitoring of batch process based on overcomplete broad learning network[J]. Engineering Applications of Artificial Intelligence, 2021, 99: 104139.
2	Qiu K P, Wang J L, Wang R T, et al. Soft sensor development based on kernel dynamic time warping and a relevant vector machine for unequal-length batch processes[J]. Expert Systems with Applications, 2021, 182: 115223.
3	王雅琳, 潘雨晴, 刘晨亮. 基于GSA-LSTM动态结构特征提取的间歇过程监测方法[J]. 化工学报, 2022, 73(9): 3994-4002.
	Wang Y L, Pan Y Q, Liu C L. Intermittent process monitoring based on GSA-LSTM dynamic structurefeature extraction[J]. CIESC Journal, 2022, 73(9): 3994-4002.
4	Zhu J L, Wang Y Q, Zhou D H, et al. Batch process modeling and monitoring with local outlier factor[J]. IEEE Transactions on Control Systems Technology, 2019, 27(4): 1552-1565.
5	Yuan X F, Huang B, Wang Y L, et al. Deep learning-based feature representation and its application for soft sensor modeling with variable-wise weighted SAE[J]. IEEE Transactions on Industrial Informatics, 2018, 14(7): 3235-3243.
6	Yamaguchi T, Yamashita Y. Quality prediction for multi-grade batch process using sparse flexible clustered multi-task learning[J]. Computers & Chemical Engineering, 2021, 150: 107320.
7	褚菲, 彭闯, 贾润达, 等. 基于多尺度核JYMKPLS迁移模型的间歇过程产品质量的在线预测方法[J]. 化工学报, 2021, 72(4): 2178-2189.
	Chu F, Peng C, Jia R D, et al. Online prediction method of batch process product quality based on multi-scale kernel JYMKPLS transfer model[J]. CIESC Journal, 2021, 72(4): 2178-2189.
8	Huo X, Hao K R, Chen L, et al. A dynamic soft sensor of industrial fuzzy time series with propositional linear temporal logic[J]. Expert Systems with Applications, 2022, 201: 117176.
9	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.
10	Yu W K, Zhao C H, Huang B. Stationary subspace analysis-based hierarchical model for batch processes monitoring[J]. IEEE Transactions on Control Systems Technology, 2020, 29(1): 444-453.
11	赵春晖, 余万科, 高福荣. 非平稳间歇过程数据解析与状态监控—回顾与展望[J]. 自动化学报, 2020, 46(10): 2072-2091.
	Zhao C H, Yu W K, Gao F R. Data analytics and condition monitoring methods for nonstationary batch processes—current status and future[J]. Acta AutomaticaSinica, 2020, 46(10): 2072-2091.
12	Qiu K P, Wang J L, Zhou X J, et al. Soft sensor framework based on semisupervised just-in-time relevance vector regression for multiphase batch processes with unlabeled data[J]. Industrial & Engineering Chemistry Research, 2020, 59(44): 19633-19642.
13	Jin H P, Chen X G, Yang J W, et al. Adaptive soft sensor modeling framework based on just-in-time learning and kernel partial least squares regression for nonlinear multiphase batch processes[J]. Computers & Chemical Engineering, 2014, 71: 77-93.
14	Zhu J L, Gao F R. Improved nonlinear quality estimation for multiphase batch processes based on relevance vector machine with neighborhood component variable selection[J]. Industrial & Engineering Chemistry Research, 2018, 57(2): 666-676.
15	Wang J L, Qiu K P, Wang R T, et al. Development of soft sensor based on sequential kernel fuzzy partitioning and just-in-time relevance vector machine for multiphase batch processes[J]. IEEE Transactions on Instrumentation and Measurement, 2021, 70: 1-10.
16	Zhao C H, Sun Y X. Step-wise sequential phase partition (SSPP) algorithm based statistical modeling and online process monitoring[J]. Chemometrics & Intelligent Laboratory Systems, 2013, 125: 109-120.
17	Zhang S M, Zhao C H, Wang S, et al. Pseudo time-slice construction using a variable moving window K nearest neighbor rule for sequential uneven phase division and batch process monitoring[J]. Industrial & Engineering Chemistry Research, 2017, 56(3): 728-740.
18	Dong W W, Yao Y, Gao F R. Phase analysis and identification method for multiphase batch processes with partitioning multi-way principal component analysis (MPCA) model[J]. Chinese Journal of Chemical Engineering, 2012, 20(6): 1121-1127.
19	Luo L J. Trajectory-based phase partition and multiphase multilinear models for monitoring and quality prediction of multiphase batch processes[J]. Journal of Chemometrics, 2018, 32(7): e3013.
20	Peng K X, Zhang K, You B, et al. A quality-based nonlinear fault diagnosis framework focusing on industrial multimode batch processes[J]. IEEE Transactions on Industrial Electronics, 2016, 63(4): 2615-2624.
21	Hui Y Y, Zhao X Q. Multi-phase batch process monitoring based on multiway weighted global neighborhood preserving embedding method[J]. Journal of Process Control, 2018, 69: 44-57.
22	Rodriguez A, Laio A. Clustering by fast search and find of density peaks[J]. Science, 2014, 344(6191): 1492-1496.
23	谭帅, 常玉清, 王福利, 等. 基于GMM的多模态过程模态识别与过程监测[J]. 控制与决策, 2015, 30(1): 53-58.
	Tan S, Chang Y Q, Wang F L, et al. Mode identification and process monitoring for multiple mode processes based on GMM[J]. Control and Decision, 2015, 30(1): 53-58.
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: 2137-2149.
25	Luo L J, Bao S Y, Mao J F, et al. Fuzzy phase partition and hybrid modeling based quality prediction and process monitoring methods for multiphase batch processes[J]. Industrial & Engineering Chemistry Research, 2016, 55(14): 4045-4058.
26	Zhang S M, Bao X L. Phase partition and online monitoring for batch processes based on Harris hawks optimization[J]. Control Engineering Practice, 2023, 138: 105554.
27	周阅昇, 熊伟丽. 基于迁移成分分析的发酵过程集成软测量建模[J]. 系统仿真学报, 2023, 35(3): 623-631.
	Zhou Y S, Xiong W L. Integrated soft sensor for fermentation process based on transfer component analysis[J]. Journal of System Simulation, 2023, 35(3): 623-631.
28	Jiang X Y, Ge Z Q. Improving the performance of just-in-time learning-based soft sensor through data augmentation[J]. IEEE Transactions on Industrial Electronics, 2022, 69(12): 13716-13726.
29	Tax D M J, Duin R P W. Support vector data description[J]. Machine Learning, 2004, 54(1): 45-66.
30	Tipping M E. Sparse Bayesian learning and the relevance vector machine[J]. Journal of machine learning research, 2001, 1(6): 211-244.
31	Birol G, Ündey C, Çinar A. A modular simulation package for fed-batch fermentation: penicillin production[J]. Computers & Chemical Engineering, 2002, 26(11): 1553-1565.
32	张雷, 张小刚, 陈华. 基于Gath-Geva算法和核极限学习机的多阶段间歇过程软测量[J]. 化工学报, 2018, 69(6): 2576-2585.
	Zhang L, Zhang X G, Chen H. Soft sensors for multi-stage batch processes based on Gath-Geva algorithm and kernel extreme learning machine[J]. CIESC Journal, 2018, 69(6): 2576-2585.

[1]	郭鑫, 李文静, 乔俊飞. 基于自组织模块化神经网络的污水处理过程出水参数预测[J]. 化工学报, 2024, 75(9): 3242-3254.
[2]	赵武灵, 满奕. 基于变分编码器的纳米纤维素分子结构预测模型框架研究[J]. 化工学报, 2024, 75(9): 3221-3230.
[3]	祝赫, 张仪, 齐娜娜, 张锴. 欧拉-欧拉双流体模型中颗粒黏性对液固散式流态化的影响[J]. 化工学报, 2024, 75(9): 3103-3112.
[4]	王倩倩, 李冰, 郑伟波, 崔国民, 赵兵涛, 明平文. 氢燃料电池局部动态特征三维模型[J]. 化工学报, 2024, 75(8): 2812-2820.
[5]	丁家琦, 刘海涛, 赵普, 朱香凝, 王晓放, 谢蓉. 煤炭超临界水制氢反应器内多相流场智能滚动预测研究[J]. 化工学报, 2024, 75(8): 2886-2896.
[6]	金虎, 杨帆, 戴梦瑶. 基于格子Boltzmann方法的液滴在圆柱壁面上运动过程研究[J]. 化工学报, 2024, 75(8): 2897-2908.
[7]	杨明军, 巩广军, 郑嘉男, 宋永臣. 泥质低渗水合物降压开采特性与模型研究[J]. 化工学报, 2024, 75(8): 2909-2916.
[8]	童永祺, 程杰, 林海, 陈曦, 赵海波. 10 MW_th化学链燃烧反应装置的CPFD模拟[J]. 化工学报, 2024, 75(8): 2949-2959.
[9]	黄静茹, 陈佳轩, 张群锋, 阮晋, 朱来, 叶光华, 周兴贵. ZSM-5分子筛结构对苯烷基化反应性能影响的数值模拟研究[J]. 化工学报, 2024, 75(7): 2544-2555.
[10]	马君霞, 李林涛, 熊伟丽. 基于Tri-training GPR的半监督软测量建模方法[J]. 化工学报, 2024, 75(7): 2613-2623.
[11]	陈彦伶, 袁炳志, 王丽伟, 张宸, 朱涵玉. 非平衡条件下金属氯化物-氨工质对的吸附动力学研究[J]. 化工学报, 2024, 75(6): 2252-2261.
[12]	黎宏陶, 王振雷, 王昕. 基于即时学习的改进条件高斯回归软测量[J]. 化工学报, 2024, 75(6): 2299-2312.
[13]	张晗, 张淑宁, 刘珂, 邓冠龙. 基于慢特征分析与最小二乘支持向量回归集成的草酸钴合成过程粒度预报[J]. 化工学报, 2024, 75(6): 2313-2321.
[14]	常成功, 宋皓楠, 雷飞霞, 狄子琛, 程芳琴. 高炉喷吹重整焦炉气工艺分析及减碳潜力研究[J]. 化工学报, 2024, 75(6): 2344-2352.
[15]	师毓辉, 邢继远, 姜雪晗, 叶爽, 黄伟光. 基于PBM的离心式叶轮内气泡破碎合并数值模拟[J]. 化工学报, 2024, 75(5): 1816-1829.