Adaptive neuro-fuzzy prediction of carbon monoxide emission from a clinical waste incineration plant

I. Norhayati, M. Rashid

Research output: Research - peer-reviewArticle

Abstract

The current practice of monitoring air emission from an incineration plant is through a hardware system known as Continuous Emission Monitoring Systems (CEMSs). Considering that CEMS suffers from high installation and maintenance cost, thus, the present work focuses on a modelling technique through an Adaptive Neuro-Fuzzy Inference System (ANFIS) to develop a predictive model of carbon monoxide (CO) emission utilizing actual data taken from a clinical waste incineration plant having capacity to process 250 kg waste/h. An hourly averaged of 1000 data points consisted of nine input–output data pairs was utilized to develop a Sugeno-type fuzzy structure by applying subtractive clustering method. As the data were divided into three sets, i.e. 70% for training, 15% for checking and the rest for testing, the values of the coefficient of determination (R2), root-mean-square error (RMSE), mean bias error (MBE) and accuracy (Ac) were calculated for each set to demonstrate its applicability and validity, emphasizing on the testing set since unseen data were exposed to the model. Result showed that ANFIS was able to learn from these data and excellently predicted the CO emission with R2, RMSE, MBE and Ac of 0.98, 4.45, 0.66 ppm and 87.98% in the testing set, respectively.

LanguageEnglish
Pages1-13
Number of pages13
JournalNeural Computing and Applications
DOIs
StateAccepted/In press - 25 Feb 2017

Fingerprint

Waste incineration
Carbon monoxide
Monitoring
Testing
Fuzzy inference
Mean square error
Incineration
Hardware
Air
Costs

Keywords

  • ANFIS
  • Clinical waste
  • CO emission
  • Continuous emission monitoring system
  • Incineration
  • Model prediction

ASJC Scopus subject areas

  • Software
  • Artificial Intelligence

Cite this

@article{6be5cb53d05540bc86cbce9a4406d781,
title = "Adaptive neuro-fuzzy prediction of carbon monoxide emission from a clinical waste incineration plant",
abstract = "The current practice of monitoring air emission from an incineration plant is through a hardware system known as Continuous Emission Monitoring Systems (CEMSs). Considering that CEMS suffers from high installation and maintenance cost, thus, the present work focuses on a modelling technique through an Adaptive Neuro-Fuzzy Inference System (ANFIS) to develop a predictive model of carbon monoxide (CO) emission utilizing actual data taken from a clinical waste incineration plant having capacity to process 250 kg waste/h. An hourly averaged of 1000 data points consisted of nine input–output data pairs was utilized to develop a Sugeno-type fuzzy structure by applying subtractive clustering method. As the data were divided into three sets, i.e. 70% for training, 15% for checking and the rest for testing, the values of the coefficient of determination (R2), root-mean-square error (RMSE), mean bias error (MBE) and accuracy (Ac) were calculated for each set to demonstrate its applicability and validity, emphasizing on the testing set since unseen data were exposed to the model. Result showed that ANFIS was able to learn from these data and excellently predicted the CO emission with R2, RMSE, MBE and Ac of 0.98, 4.45, 0.66 ppm and 87.98% in the testing set, respectively.",
keywords = "ANFIS, Clinical waste, CO emission, Continuous emission monitoring system, Incineration, Model prediction",
author = "I. Norhayati and M. Rashid",
year = "2017",
month = "2",
doi = "10.1007/s00521-017-2921-z",
pages = "1--13",
journal = "Neural Computing and Applications",
issn = "0941-0643",
publisher = "Springer London",

}

TY - JOUR

T1 - Adaptive neuro-fuzzy prediction of carbon monoxide emission from a clinical waste incineration plant

AU - Norhayati,I.

AU - Rashid,M.

PY - 2017/2/25

Y1 - 2017/2/25

N2 - The current practice of monitoring air emission from an incineration plant is through a hardware system known as Continuous Emission Monitoring Systems (CEMSs). Considering that CEMS suffers from high installation and maintenance cost, thus, the present work focuses on a modelling technique through an Adaptive Neuro-Fuzzy Inference System (ANFIS) to develop a predictive model of carbon monoxide (CO) emission utilizing actual data taken from a clinical waste incineration plant having capacity to process 250 kg waste/h. An hourly averaged of 1000 data points consisted of nine input–output data pairs was utilized to develop a Sugeno-type fuzzy structure by applying subtractive clustering method. As the data were divided into three sets, i.e. 70% for training, 15% for checking and the rest for testing, the values of the coefficient of determination (R2), root-mean-square error (RMSE), mean bias error (MBE) and accuracy (Ac) were calculated for each set to demonstrate its applicability and validity, emphasizing on the testing set since unseen data were exposed to the model. Result showed that ANFIS was able to learn from these data and excellently predicted the CO emission with R2, RMSE, MBE and Ac of 0.98, 4.45, 0.66 ppm and 87.98% in the testing set, respectively.

AB - The current practice of monitoring air emission from an incineration plant is through a hardware system known as Continuous Emission Monitoring Systems (CEMSs). Considering that CEMS suffers from high installation and maintenance cost, thus, the present work focuses on a modelling technique through an Adaptive Neuro-Fuzzy Inference System (ANFIS) to develop a predictive model of carbon monoxide (CO) emission utilizing actual data taken from a clinical waste incineration plant having capacity to process 250 kg waste/h. An hourly averaged of 1000 data points consisted of nine input–output data pairs was utilized to develop a Sugeno-type fuzzy structure by applying subtractive clustering method. As the data were divided into three sets, i.e. 70% for training, 15% for checking and the rest for testing, the values of the coefficient of determination (R2), root-mean-square error (RMSE), mean bias error (MBE) and accuracy (Ac) were calculated for each set to demonstrate its applicability and validity, emphasizing on the testing set since unseen data were exposed to the model. Result showed that ANFIS was able to learn from these data and excellently predicted the CO emission with R2, RMSE, MBE and Ac of 0.98, 4.45, 0.66 ppm and 87.98% in the testing set, respectively.

KW - ANFIS

KW - Clinical waste

KW - CO emission

KW - Continuous emission monitoring system

KW - Incineration

KW - Model prediction

UR - http://www.scopus.com/inward/record.url?scp=85013854135&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85013854135&partnerID=8YFLogxK

U2 - 10.1007/s00521-017-2921-z

DO - 10.1007/s00521-017-2921-z

M3 - Article

SP - 1

EP - 13

JO - Neural Computing and Applications

T2 - Neural Computing and Applications

JF - Neural Computing and Applications

SN - 0941-0643

ER -