Dynamic data-driven prediction of lean blowout in a swirl-stabilized combustor

Soumalya Sarkar, Asok Ray, Achintya Mukhopadhyay, Swarnendu Sen

Research output: Contribution to journalArticle

14 Citations (Scopus)

Abstract

This paper addresses dynamic data-driven prediction of lean blowout (LBO) phenomena in confined combustion processes, which are prevalent in many physical applications (e.g., landbased and aircraft gas-turbine engines). The underlying concept is built upon pattern classification and is validated for LBO prediction with time series of chemiluminescence sensor data from a laboratory-scale swirl-stabilized dump combustor. The proposed method of LBO prediction makes use of the theory of symbolic dynamics, where (finite-length) time series data are partitioned to produce symbol strings that, in turn, generate a special class of probabilistic finite state automata (PFSA). These PFSA, called D-Markov machines, have a deterministic algebraic structure and their states are represented by symbol blocks of length D or less, where D is a positive integer. The D-Markov machines are constructed in two steps: (i) state splitting, i.e., the states are split based on their information contents, and (ii) state merging, i.e., two or more states (of possibly different lengths) are merged together to form a new state without any significant loss of the embedded information. The modeling complexity (e.g., number of states) of a D-Markov machine model is observed to be drastically reduced as the combustor approaches LBO. An anomaly measure, based on Kullback-Leibler divergence, is constructed to predict the proximity of LBO. The problem of LBO prediction is posed in a pattern classification setting and the underlying algorithms have been tested on experimental data at different extents of fuel-air premixing and fuel/air ratio. It is shown that, over a wide range of fuel-air premixing, D-Markov machines with D > 1 perform better as predictors of LBO than those with D = 1.

Original languageEnglish (US)
Pages (from-to)209-242
Number of pages34
JournalInternational Journal of Spray and Combustion Dynamics
Volume7
Issue number3
DOIs
StatePublished - Sep 2015

Fingerprint

combustion chambers
Combustors
premixing
Finite automata
predictions
Pattern recognition
Time series
Air
dump combustors
Chemiluminescence
fuel-air ratio
gas turbine engines
Merging
air
chemiluminescence
Gas turbines
Turbines
Aircraft
aircraft
integers

All Science Journal Classification (ASJC) codes

  • Automotive Engineering
  • Energy Engineering and Power Technology
  • Physics and Astronomy(all)

Cite this

Sarkar, Soumalya ; Ray, Asok ; Mukhopadhyay, Achintya ; Sen, Swarnendu. / Dynamic data-driven prediction of lean blowout in a swirl-stabilized combustor. In: International Journal of Spray and Combustion Dynamics. 2015 ; Vol. 7, No. 3. pp. 209-242.
@article{78a500ac8295418880f0b01fef3ce351,
title = "Dynamic data-driven prediction of lean blowout in a swirl-stabilized combustor",
abstract = "This paper addresses dynamic data-driven prediction of lean blowout (LBO) phenomena in confined combustion processes, which are prevalent in many physical applications (e.g., landbased and aircraft gas-turbine engines). The underlying concept is built upon pattern classification and is validated for LBO prediction with time series of chemiluminescence sensor data from a laboratory-scale swirl-stabilized dump combustor. The proposed method of LBO prediction makes use of the theory of symbolic dynamics, where (finite-length) time series data are partitioned to produce symbol strings that, in turn, generate a special class of probabilistic finite state automata (PFSA). These PFSA, called D-Markov machines, have a deterministic algebraic structure and their states are represented by symbol blocks of length D or less, where D is a positive integer. The D-Markov machines are constructed in two steps: (i) state splitting, i.e., the states are split based on their information contents, and (ii) state merging, i.e., two or more states (of possibly different lengths) are merged together to form a new state without any significant loss of the embedded information. The modeling complexity (e.g., number of states) of a D-Markov machine model is observed to be drastically reduced as the combustor approaches LBO. An anomaly measure, based on Kullback-Leibler divergence, is constructed to predict the proximity of LBO. The problem of LBO prediction is posed in a pattern classification setting and the underlying algorithms have been tested on experimental data at different extents of fuel-air premixing and fuel/air ratio. It is shown that, over a wide range of fuel-air premixing, D-Markov machines with D > 1 perform better as predictors of LBO than those with D = 1.",
author = "Soumalya Sarkar and Asok Ray and Achintya Mukhopadhyay and Swarnendu Sen",
year = "2015",
month = "9",
doi = "10.1260/1756-8277.7.3.209",
language = "English (US)",
volume = "7",
pages = "209--242",
journal = "International Journal of Spray and Combustion Dynamics",
issn = "1756-8277",
publisher = "Multi-Science Publishing Co. Ltd",
number = "3",

}

Dynamic data-driven prediction of lean blowout in a swirl-stabilized combustor. / Sarkar, Soumalya; Ray, Asok; Mukhopadhyay, Achintya; Sen, Swarnendu.

In: International Journal of Spray and Combustion Dynamics, Vol. 7, No. 3, 09.2015, p. 209-242.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Dynamic data-driven prediction of lean blowout in a swirl-stabilized combustor

AU - Sarkar, Soumalya

AU - Ray, Asok

AU - Mukhopadhyay, Achintya

AU - Sen, Swarnendu

PY - 2015/9

Y1 - 2015/9

N2 - This paper addresses dynamic data-driven prediction of lean blowout (LBO) phenomena in confined combustion processes, which are prevalent in many physical applications (e.g., landbased and aircraft gas-turbine engines). The underlying concept is built upon pattern classification and is validated for LBO prediction with time series of chemiluminescence sensor data from a laboratory-scale swirl-stabilized dump combustor. The proposed method of LBO prediction makes use of the theory of symbolic dynamics, where (finite-length) time series data are partitioned to produce symbol strings that, in turn, generate a special class of probabilistic finite state automata (PFSA). These PFSA, called D-Markov machines, have a deterministic algebraic structure and their states are represented by symbol blocks of length D or less, where D is a positive integer. The D-Markov machines are constructed in two steps: (i) state splitting, i.e., the states are split based on their information contents, and (ii) state merging, i.e., two or more states (of possibly different lengths) are merged together to form a new state without any significant loss of the embedded information. The modeling complexity (e.g., number of states) of a D-Markov machine model is observed to be drastically reduced as the combustor approaches LBO. An anomaly measure, based on Kullback-Leibler divergence, is constructed to predict the proximity of LBO. The problem of LBO prediction is posed in a pattern classification setting and the underlying algorithms have been tested on experimental data at different extents of fuel-air premixing and fuel/air ratio. It is shown that, over a wide range of fuel-air premixing, D-Markov machines with D > 1 perform better as predictors of LBO than those with D = 1.

AB - This paper addresses dynamic data-driven prediction of lean blowout (LBO) phenomena in confined combustion processes, which are prevalent in many physical applications (e.g., landbased and aircraft gas-turbine engines). The underlying concept is built upon pattern classification and is validated for LBO prediction with time series of chemiluminescence sensor data from a laboratory-scale swirl-stabilized dump combustor. The proposed method of LBO prediction makes use of the theory of symbolic dynamics, where (finite-length) time series data are partitioned to produce symbol strings that, in turn, generate a special class of probabilistic finite state automata (PFSA). These PFSA, called D-Markov machines, have a deterministic algebraic structure and their states are represented by symbol blocks of length D or less, where D is a positive integer. The D-Markov machines are constructed in two steps: (i) state splitting, i.e., the states are split based on their information contents, and (ii) state merging, i.e., two or more states (of possibly different lengths) are merged together to form a new state without any significant loss of the embedded information. The modeling complexity (e.g., number of states) of a D-Markov machine model is observed to be drastically reduced as the combustor approaches LBO. An anomaly measure, based on Kullback-Leibler divergence, is constructed to predict the proximity of LBO. The problem of LBO prediction is posed in a pattern classification setting and the underlying algorithms have been tested on experimental data at different extents of fuel-air premixing and fuel/air ratio. It is shown that, over a wide range of fuel-air premixing, D-Markov machines with D > 1 perform better as predictors of LBO than those with D = 1.

UR - http://www.scopus.com/inward/record.url?scp=84959512784&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84959512784&partnerID=8YFLogxK

U2 - 10.1260/1756-8277.7.3.209

DO - 10.1260/1756-8277.7.3.209

M3 - Article

AN - SCOPUS:84959512784

VL - 7

SP - 209

EP - 242

JO - International Journal of Spray and Combustion Dynamics

JF - International Journal of Spray and Combustion Dynamics

SN - 1756-8277

IS - 3

ER -