基于递推子空间的机组数字孪生模型预测精度优化方法

doi:10.16182/j.issn1004731x.joss.22-0989

摘要/Abstract

摘要：

由于机理分析的简化假设条件或设备实际运行中参数特性偏移等因素，导致机理建模不可避免存在模型误差。针对该问题，提出一种基于递推子空间的火电机组数字孪生模型预测精度优化方法。分析机组关键设备的运行机制，结合典型工况小样本数据，建立火电机组的全设备多工况非线性动态机理模型，确保数字孪生系统模型具有较好的可解释性与泛化性能；基于递推子空间辨识方法，建立预测精度优化模型并实时进行在线更新，补偿机理模型产生的误差，提高整体模型的预测精度，保证数字孪生模型的高保真性。仿真实验验证了所提方法的有效性。

关键词: 数字孪生, 火电机组, 模型预测精度优化, 子空间辨识, 数据驱动

Abstract:

Due to factors such as simplified assumptions or equipment characteristic deviation, modeling errors are inevitable in the mechanism modeling of thermal power units. To deal with the problem, this paper proposes a novel model refinement method based on recursive subspace for the digital twin of thermal power units. Firstly, the digital twin models are built based on mechanism analysis and combined with small sample data of typical conditions, ensuring interpretability and generalization performance. Secondly, based on the recursive subspace identification method, the refinement model is built and updated online in real time to compensate for the modeling error, improving the prediction accuracy and ensuring the high fidelity of the overall digital twin model. Finally, simulation results validate the effectiveness of the proposed method.

Key words: digital twin, thermal power unit, model refinement, subspace identification, data driven

中图分类号:

TP391.9

赵彦博,蔡远利,胡怀中 . 基于递推子空间的机组数字孪生模型预测精度优化方法[J]. 系统仿真学报, 2024, 36(1): 195-202.

Zhao Yanbo,Cai Yuanli,Hu Huaizhong . Recursive Subspace-based Model Refinement Method for Digital Twin of Thermal Power Unit[J]. Journal of System Simulation, 2024, 36(1): 195-202.

图/表 14

表1

变量及参数释义

变量	释义
$W g$	给煤量
$μ h p c$	主蒸汽阀门开度
$P h p c$	主蒸汽压力
$P w$	输出功率
$T w$	水冷壁惯性时间常数
$T B$	锅炉惯性时间常数
$K D$	汽包比例系数
$K u$	汽包到过热器泄流系数
$K B$	锅炉比例系数
$T D$	汽包惯性时间常数
$K H$	高压缸比例增益
$K I$	中压缸比例增益
$K L$	低压缸比例增益
$T H$	高压缸惯性时间常数
$T I$	中压缸惯性时间常数
$T L$	低压缸惯性时间常数
$K q$	锅炉入口热量修正系数
$Q n e t, a r$	燃煤低位发热量

表1

图1

图2

图3

表2

表3

图4

图5

表4

表5

图6

图7

表6

表7

参考文献 18

1	de Mello F P. Boiler Models for System Dynamic Performance Studies[J]. IEEE Transactions on Power Systems, 1991, 6(1): 66-74.
2	Grieves M, Vickers J. Digital Twin: Mitigating Unpredictable, Undesirable Emergent Behavior in Complex Systems[M]. Cham: Springer International Publishing, 2017: 85-113.
3	陶飞, 马昕, 胡天亮, 等. 数字孪生标准体系[J]. 计算机集成制造系统, 2019, 25(10): 2405-2418.
	Tao Fei, Ma Xin, Hu Tianliang, et al. Research on Digital Twin Standard System[J]. Computer Integrated Manufacturing Systems, 2019, 25(10): 2405-2418.
4	Zhang Lin, Zhou Longfei, Horn B K P. Building a Right Digital Twin with Model Engineering[J]. Journal of Manufacturing Systems, 2021, 59: 151-164.
5	蔡远利, 高鑫, 张渊. 数字孪生技术的概念、方法及应用[C]//第二十届中国系统仿真技术及其应用学术年会论文集. 合肥: 中国科学技术大学出版社, 2019: 141-145.
6	李胜男, 谭鹏, 饶德备, 等. 融合数据与机理的燃煤发电机组协调系统建模[J]. 工程热物理学报, 2022, 43(1): 19-26.
	Li Shengnan, Tan Peng, Rao Debei, et al. Modeling of Coal-fired Generating Unit Coordination System with Data and Mechanism[J]. Journal of Engineering Thermophysics, 2022, 43(1): 19-26.
7	王占能, 徐祖华, 赵均, 等. 基于负荷划分数据和支持向量机的火电厂燃烧过程建模[J]. 化工学报, 2013, 64(12): 4496-4502.
	Wang Zhanneng, Xu Zuhua, Zhao Jun, et al. Coal-fired Power Plant Boiler Combustion Process Modeling Based on Support Vector Machine and Load Data Division[J]. CIESC Journal, 2013, 64(12): 4496-4502.
8	Liu X J, Kong X B, Hou G L, et al. Modeling of a 1000MW Power Plant Ultra Super-critical Boiler System Using Fuzzy-neural Network Methods[J]. Energy Conversion and Management, 2013, 65: 518-527.
9	尹二新, 张同卫, 唐坚, 等. 火电机组过程大数据建模方法研究[J]. 电力大数据, 2019, 22(10): 30-35.
	Yin Erxin, Zhang Tongwei, Tang Jian, et al. Research on Industrial Process Big Data Modeling Method for Thermal Power Unit[J]. Power Systems and Big Data, 2019, 22(10): 30-35.
10	Qin S J. An Overview of Subspace Identification[J]. Computers & Chemical Engineering, 2006, 30(10/12): 1502-1513.
11	Kadali Ramesh, Huang Biao, Rossiter Anthony. A Data Driven Subspace Approach to Predictive Controller Design[J]. Control Engineering Practice, 2003, 11(3): 261-278.
12	Pepijn Bastiaan Cox, Tóth Roland. Linear Parameter-varying Subspace Identification: a Unified Framework[J]. Automatica, 2021, 123: 109296.
13	Ghosh Debanjan, Hermonat Emma, Mhaskar Prashant, et al. Hybrid Modeling Approach Integrating First-principles Models with Subspace Identification[J]. Industrial & Engineering Chemistry Research, 2019, 58(30): 13533-13543.
14	罗小锁, 周国清, 邹涛. 基于子空间辨识的状态空间模型预测控制[J]. 计算机工程与应用, 2012, 48(19): 234-237, 248.
	Luo Xiaosuo, Zhou Guoqing, Zou Tao. State Space Model Predictive Control Based on Subspace Identification[J]. Computer Engineering and Applications, 2012, 48(19): 234-237, 248.
15	褚德海, 张婷, 董泽. 基于闭环子空间辨识的电厂协调系统预测控制[J]. 计算机仿真, 2019, 36(5): 128-132, 174.
	Chu Dehai, Zhang Ting, Dong Ze. Predictive Control for a Coordinated Control System of a Power Plant Based on Closed-loop Subspace Identification[J]. Computer Simulation, 2019, 36(5): 128-132, 174.
16	罗小锁, 丁宝苍, 邹涛. 基于在线子空间辨识的自适应预测控制[J]. 化工自动化及仪表, 2010, 37(10): 6-9.
	Luo Xiaosuo, Ding Baocang, Zou Tao. Adaptive Predictive Control Based on On-line Subspace Identification[J]. Control and Instruments in Chemical Industry, 2010, 37(10): 6-9.
17	谢磊, 梁武星, 张泉灵, 等. 基于快速滑窗QR分解的自适应子空间辨识[J]. 化工学报, 2008, 59(6): 1448-1453.
	Xie Lei, Liang Wuxing, Zhang Quanling, et al. Adaptive Subspace Identification Based on Fast Moving Window QR Decomposition[J]. Journal of Chemical Industry and Engineering(China), 2008, 59(6): 1448-1453.
18	苏奇全, 贾宏光, 朱明超, 等. 基于递推闭环子空间辨识的自适应预测控制方法[J]. 信息与控制, 2015, 44(2): 252-256.
	Su Qiquan, Jia Hongguang, Zhu Mingchao, et al. An Adaptive Predictive Control Method Based on Recursive Closed-loop Subspace Identification[J]. Information and Control, 2015, 44(2): 252-256.

EI	主蒸汽压力(MPa)				误差降幅(%)
EI	NOMR	SMR	ANNMR	RSMR	SMR	ANNMR	RSMR
EA	0.004 0	0.000 7	0.000 6	0.000 1	82.5	85.0	97.6
EMAX	0.004 7	0.001 3	0.001 4	0.000 3	72.3	70.2	93.6

EI	输出功率/MW				误差降幅/%
EI	NOMR	SMR	ANNMR	RSMR	SMR	ANNMR	RSMR
EA	4.722 5	0.678 4	0.740 7	0.402 8	85.6	84.3	91.5
EMAX	5.186 2	0.950 2	1.211 0	0.532 0	81.7	76.7	89.7

EI	主蒸汽压力/MPa				误差降幅/%
EI	NOMR	SMR	ANNMR	RSMR	SMR	ANNMR	RSMR
EA	0.004 0	0.001 1	0.000 7	0.000 3	72.5	82.5	92.9
EMAX	0.005 5	0.002 5	0.001 6	0.000 7	54.6	70.9	87.3

EI	输出功率/MW				误差降幅/%
EI	NOMR	SMR	ANNMR	RSMR	SMR	ANNMR	RSMR
EA	5.278 8	0.749 3	0.263 0	0.137 2	85.8	95.0	97.4
EMAX	5.686 7	0.806 7	0.679 3	0.192 8	85.8	88.1	96.6

EI	主蒸汽压力/MPa				误差降幅/%
EI	NOMR	SMR	ANNMR	RSMR	SMR	ANNMR	RSMR
EA	0.004 2	0.001 6	0.001 6	0.000 1	61.9	61.9	97.6
EMAX	0.008 1	0.002 2	0.002 5	0.000 4	72.8	69.1	95.1