Artificial neural network-based estimation of mercury speciation in combustion flue gases

Jensen, Robert; Karki, Shankar; Salehfar, Hossein

doi:10.1016/j.fuproc.2003.11.020

Cited by 41 publications

(29 citation statements)

References 2 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…2 indicate the correlation between the actual and predicted values for the three types of mercury speciation when the optimum abductive models derived from Tables 2-4 are evaluated on the full training set of 72 samples in each case. The correlation coefficients range from 0.932 to 0.972 which compare favorably with the value of 0.987 quoted by [13] for a much smaller (13-sample) randomly selected subset of training data. Performance of the three optimum models on the external evaluation set of 10 samples is shown in Fig.…”

Section: Single Abductive Network Models For Mercury Speciationsupporting

confidence: 73%

“…Therefore, obtained evaluation results should give a good indication of model performance with new cases under similar prevailing conditions. It would be interesting to see how the abductive network models perform on data for power plants in the ICR scheme other than the 82 plants for which data was given in [13]. As a performance metric, we used the mean absolute error (MAE) defined as the average of the absolute values of the error between the actual and predicted values taken over all the samples considered.…”

Section: Single Abductive Network Models For Mercury Speciationmentioning

confidence: 99%

“…Applications of neural networks to modeling mercury speciation in flue gases are relatively scarce in the literature, and only the work by Jensen et al [13] could be cited at the time of writing.…”

Section: Introductionmentioning

confidence: 99%

“…We used a dataset of experimental measurements compiled by Jensen et al [13]. The dataset consists of 82 samples of measurements, each representing a coal-fired power plant in the USA as part of the Information Collection Request (ICR) scheme run by the Environmental Protection Agency (EPA) in 1998 [13].…”

Section: Introductionmentioning

confidence: 99%

“…The dataset consists of 82 samples of measurements, each representing a coal-fired power plant in the USA as part of the Information Collection Request (ICR) scheme run by the Environmental Protection Agency (EPA) in 1998 [13]. Each record contains six explanatory input variables representing the coal properties and the boiler operating conditions that are known to affect mercury speciation emission, as well as three outputs representing elemental, oxidized, and particulate mercury in the flue gas of the boiler.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Predictive modeling of mercury speciation in combustion flue gases using GMDH-based abductive networks

Abdel-Aal

2007

Fuel Processing Technology

View full text Add to dashboard Cite

Abstract:Modeling mercury speciation is an important requirement for estimating harmful emissions from coal-fired power plants and developing strategies to reduce them. First principle models based on chemical, kinetic, and thermodynamic aspects exist, but these are complex and difficult to develop. The use of modern data-based machine learning techniques has been recently introduced, including neural networks. Here we propose an alternative approach using abductive networks based on the group method of data handling (GMDH) algorithm, with the advantages of simplified and more automated model synthesis, automatic selection of significant inputs, and more transparent input-output model relationships. Models were developed for predicting three types of mercury speciation (elemental, oxidized, and particulate) using a small data set containing six inputs parameters on the composition of the coal used and boiler operating conditions. Prediction performance compares favourably with neural network models developed using the same dataset, with correlation coefficients as high as 0.97 for training data. Network committees (ensembles) are proposed as a means of improving prediction accuracy, and suggestions are made for future work to further improve performance.

show abstract

Section: Single Abductive Network Models For Mercury Speciationsupporting

confidence: 73%

Section: Single Abductive Network Models For Mercury Speciationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations