Identification for Wiener‐Hammerstein systems under quantized inputs and quantized output observations

Guo, Jin; Zhao, Yanlong

doi:10.1002/asjc.2237

Cited by 13 publications

(9 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…2a Then, we can relabel these states so that the corresponding transition frequency matrix has a block diagonal form. We rearrange these states by mapping {s 3…”

Section: Fast Group Separation Methodmentioning

confidence: 99%

“…One of the main difficulties is that the computational results depend sensitively on the network structure, which we refer to as a robustness issue. For some recent work on system identification of non-linear systems, systems with quantised observations, and estimation of hyperparameters, we refer the reader to [2][3][4][5] and references therein.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Two‐time scale reinforcement learning and applications to production planning

Zhang

Yin

Wang

2020

IET Control Theory & Appl

View full text Add to dashboard Cite

This study is concerned with reinforcement learning enhanced by two-time scale approximations. Many systems arising in applications are large and complex. To treat these problems, it is often beneficial, and sometimes necessary, to reduce the dimensionality and aggregate states that are 'close' to each other. In this study, the authors propose a two-time scale reinforcement learning method for such an aggregation process. In particular, they present how to classify states that are 'close' and demonstrate the effectiveness of the authors' state aggregation based two-time scale methods. Thus the problem can be considered as using learning for identifying the system. A production planning problem with failure-prone machines is used throughout this study to illustrate the main ideas, key steps and results. Monte Carlo simulations are used to generate the random environment.

show abstract

“…2a Then, we can relabel these states so that the corresponding transition frequency matrix has a block diagonal form. We rearrange these states by mapping {s 3…”

Section: Fast Group Separation Methodmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Two‐time scale reinforcement learning and applications to production planning

Zhang

Yin

Wang

2020

IET Control Theory & Appl

View full text Add to dashboard Cite

show abstract

“…In addition, the two linear blocks are considered to be parametric. Some studies require a specially designed input signal 21–23 …”

Section: Introductionmentioning

confidence: 99%

Wiener–Hammerstein nonlinear system identification using spectral analysis

Brouri

2022

Intl J Robust & Nonlinear

View full text Add to dashboard Cite

In this paper, a new approach is developed for the identification of Wiener-Hammerstein model structures. Unlike several other papers, the transfer functions of linear elements are of unknown structure. Indeed, the linear subsystems are not necessarily parametric or FIR. Furthermore, the function of nonlinear block can be nonparametric and noninvertible. The main challenge in identifying a Wiener-Hammerstein system lies in the separation of the dynamics over the input and the output LTI elements. Presently, this issue does not arise, without any conditions being imposed on the frequency band. Then, a two-stage identification method is suggested. Firstly, the system is excited by a set of constant inputs to capture the system nonlinearity. In the second stage, an identification approach based on the spectral analysis and using periodic input signals is developed to determine the linear elements parameters. In the present method, very interesting concepts are used, like the Fourier decomposition, the frequency approach and the spectrum analysis. Simulations demonstrate that the proposed identification method is both effective and efficient.

show abstract

“…For Wiener-Hammerstein models, the identification procedure generally begins with the estimation of the linear dynamics, and the main challenge is to split the LTI blocks in two. A non-exhaustive list of options includes, the "brute-force" method and the advanced method [12]; the scanning technique based on the estimation of the quadratic/cubic best linear approximation (QBLA/CBLA) [13]; the non-parametric approach based on the estimation of the QBLA [14]; fractional approach [15]; WH-EA: An Evolutionary Algorithm for Wiener-Hammerstein System Identification [16]; multistage identification [17], [18]; classification of poles and zeros using QBLA and manual tuning [19]; among others [20]- [23].…”

Section: Introductionmentioning

confidence: 99%