Identification of Slow-Rate Integrated Measurement Systems Using Expectation-Maximization Algorithm

Despite the fact that slow-rate integrated measurement (SRTM), a gradual accumulation of material during a period of time, is a well-known method of sampling in industrial systems, especially chemical processes, the problem of the identification of SRTM systems has not been addressed so far. In this...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on instrumentation and measurement Vol. 69; no. 12; pp. 9477 - 9484
Main Authors: Kheirandish, Amid, Fatehi, Alireza, Gheibi, Mir Sajjad
Format: Journal Article
Language:English
Published: New York IEEE 01.12.2020
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Subjects:
Algorithms
Autoregressive models
Autoregressive processes
Chemical reactions
Computer simulation
Expectation-maximization algorithms
Expectation–maximization (EM) algorithm
Impulse response
integrated autoregressive exogenous (IARX) model
integrated finite impulse response (IFIR) model
Kalman filters
Mathematical model
Maximization
Monte Carlo simulation
Optimization
Parameter estimation
Sampling
slow-rate integrated measurement (SRTM)
System identification
ISSN:0018-9456, 1557-9662
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Despite the fact that slow-rate integrated measurement (SRTM), a gradual accumulation of material during a period of time, is a well-known method of sampling in industrial systems, especially chemical processes, the problem of the identification of SRTM systems has not been addressed so far. In this regard, system identification of the processes having SRTM will be presented in this work. By selecting finite impulse response (FIR) and autoregressive exogenous (ARX) models for the systems, parameters of them will be accurately estimated in the framework of the expectation-maximization (EM) algorithm. For this purpose, the instantaneous values of the output at the fast-rate sampling time are assumed to be latent variables. Applying the EM algorithm leads to some high-dimensional integrals, for which Monte Carlo simulation is adopted. The effectiveness of the proposed method is illustrated by both a simulation study and implementation on a laboratory-scale pH neutralization pilot plant.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0018-9456
1557-9662
DOI:10.1109/TIM.2020.3006664