Recursive subspace identification based on random distribution theory

被引：0

作者：

Yu M. ^{[1
]}

Liu J.-C. ^{[2
,3
]}

Guo G. ^{[1
,3
]}

机构：

[1] School of Control Engineering, Northeastern University at Qinhuangdao, Qinhuangdao

[2] College of Information Science and Engineering, Northeastern University, Shenyang

[3] State Key Laboratory of Synthetical Automation for Process Industries, Northeastern University, Shenyang

来源：

Kongzhi Lilun Yu Yingyong/Control Theory and Applications | 2021年 / 38卷 / 09期

基金：

中国国家自然科学基金;

关键词：

Continuous-time systems; Online identification; Random distribution theory; Recursive subspace identification;

D O I：

10.7641/CTA.2021.00819

中图分类号：

学科分类号：

摘要：

In the practical process, the offline model estimated by the traditional subspace identification method cannot track the dynamic of the systems effectively. Although the numerical robustness of the algorithm is improved by the singular value decomposition, it also increases the difficulty of online recursion in subspace identification process. In order to solve the problems above, this paper presents a recursive subspace identification method for continuous-time stochastic systems via random distribution theory. Firstly, the continuous random distribution function is deduced by random distribution theory and the input-output matrix equation of the systems is obtained by the differential calculation. Secondly, we reduce the computational burden and storage cost by keeping the size of input and output data to be constant. Finally, the system matrices and noise intensity are updated recursively by the least square method and residual analysis. Simulation results show the efficiency and accuracy of the proposed method. © 2021, Editorial Department of Control Theory & Applications South China University of Technology. All right reserved.

引用

页码：1333 / 1340

页数：7

共 21 条

[1]

LIU T, HOU J, QIN S J, Et al., Subspace model identification under load disturbance with unknowntransient and periodic dynamics, Journal of Process Control, 85, pp. 100-111, (2020)

[2]

SINQUIN B, VERHAEGEN M., K4SID: Large-scale subspace identification with kronecker modeling, IEEE Transactions on Automatic Control, 64, 3, pp. 960-975, (2019)

[3]

HU Y S, JIANG Y F, CALLAFON R A D., Variance reduction in covariance based realization algorithm with application to closed-loop data, Automatica, 113, (2020)

[4]

HOUTZAGER I, WINGERDEN J W V, VERHAEGEN M., Recursive predictor-based subspace identification with application to the real-time closed-loop tracking of flutter, IEEE Transactions on Automatic Control, 20, 4, pp. 934-949, (2012)

[5]

LIU T, HUANG B, QIN S J., Bias-eliminated subspace model identification under time-varying deterministic type load disturbance, Journal of Process Control, 25, pp. 41-49, (2015)

[6]

OHSUMI A, KAWANO T., Subspace identification for a class of timevarying continuous-time stochastic systems via distribution-based approach, The 15th Triennial World Congress, pp. 241-246, (2002)

[7]

YU M, LIU J C., Nuclear norm-based recursive subspace prediction of time-varying continuous-time stochastic systems via distribution theory, Journal of the Franklin Institute, 355, pp. 8830-8856, (2018)

[8]

KAMEYAMA K, OHSUMI A., Subspace-based prediction of linear time-varying stochastic systems, Automatica, 43, pp. 2009-2021, (2007)

[9]

SHANG L L, LIU J C, TURKSON K., Stable recursive canonical variate state space modeling for time-varying processes, Control Engineering Practice, 36, pp. 113-119, (2015)

[10]

ZHOU P, DAI P, SONG H, Et al., Data-driven recursive subspace identification based online modelling for prediction and control of molten iron quality in blast furnace ironmaking, IET Control Theory & Applications, 11, 14, pp. 2343-2351, (2017)

← 1 2 3 →