A Feasibility Study of a Predictive Emissions Monitoring System Applied to Taipower’s Nanpu and Hsinta Power Plants

Chien, Tsung Wen; Hsueh, Hsin Ta; Chu, Hsin; Hsu, Wen Dung; Tu, Yueh-Yuan; Tsai, Hsien Shiou; Chen, Kuo-Yi

doi:10.3155/1047-3289.60.8.907

Cited by 8 publications

(3 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In 2010, Chien et al applied the same model to the same size and type cogeneration unit but at a different power station, Nanpo station, and concluded that the statistical model could not be transferred from one piece of equipment to another due to differences in equipment condition, operating model, maintenance and other parameters. [24].…”

Section: Pems Researchmentioning

confidence: 99%

Development of Predictive Emissions Monitoring System Using Open Source Machine Learning Library – Keras: A Case Study on a Cogeneration Unit

Tarnoczi

Wiens

et al. 2019

IEEE Access

View full text Add to dashboard Cite

The study provides an overview of Predictive Emissions Monitoring System's (PEMS) research, application, installation, and regulatory framework as well as develops predictive models for NO x emissions from a natural gas fired cogeneration unit using an open source machine learning library, Keras, and open source programming languages, Python and R. Nine neural network based predictive models were trained with 12 086 examples and tested with 3020 examples. The neural network-based models use eight process parameters as inputs to predict NO x emissions. All models meet the regulatory requirements for precision. The best model (32-64-64-64) has four hidden layers and uses the Nadam method for optimization. The best model has a mean absolute error of 0.5982, r-value of 0.9451, and a difference of 0.14% between the measured and predicted emission values using the test dataset. The study demonstrated the feasibility of using open source machine learning library in PEMS development. It also provides guidance to facility operators to develop their own PEMS models for monitoring emissions.INDEX TERMS Air emissions monitoring, environmental monitoring, Keras, machine learning, NO x , PEMS, predictive emissions monitoring system, tensorflow.

show abstract

Section: Pems Researchmentioning

confidence: 99%

Development of Predictive Emissions Monitoring System Using Open Source Machine Learning Library – Keras: A Case Study on a Cogeneration Unit

Tarnoczi

Wiens

et al. 2019

IEEE Access

View full text Add to dashboard Cite

show abstract

“…Computer-based predictive emission monitoring systems (PEMSs) have become attractive because of their low capital and operating costs, and low maintenance (Chien et al, 2010;Cooper and Andreasson, 1999;Harnevie et al, 1996). PEMSs can not only be used to replace existing aging CEMSs but also can be installed on small combustion sources, for which CEMSs are normally cost-prohibitive for emission monitoring.…”

Section: Introductionmentioning

confidence: 99%

Long-term evaluation of machine learning based methods for air emission monitoring

Si,

Wiens,

2023

Preprint

View full text Add to dashboard Cite

Machine learning (ML) techniques have been researched and used in various environmental monitoring applications. Few studies have reported the long-term evaluation of such applications. Discussions regarding the risks and regulatory frameworks of ML applications in environmental monitoring have been rare. We monitored the performance of six ML-based predictive models for 28 months. The six ML models used to predict NOx emissions were developed using six different algorithms. The model developed with a moderate complexity algorithm, adaptive boosting, had the best performance in long-term monitoring, with a root mean square error (RMSE) of 0.48 kg/hr in the 28-month monitoring period, and passed two of the three relative accuracy test audits. High complexity models based on gradient boosting and neural network algorithms had the best training performance, with a minimum RMSE of 0.23 kg/hr and 0.26 kg/hr, but also had the worst RMSE scores, of 0.51 kg/hr and 0.57 kg/hr, during the monitoring period. In addition, all six models failed all three relative accuracy test audits. The following problems were observed: 1) Complex ML models tended to have overfitting problems, thus indicating the importance of the trade-off between model accuracy and complexity. 2) Model input sensor drift or out of high-frequency ranges from the training data resulted in inaccurate predictions or an accuracy lower than the minimum allowed by regulators. 3) Existing regulatory frameworks must be modernized to keep pace with current machine learning practices. Some statistical tests are unsuitable for applications developed by using ML methods.

show abstract

“…At present, almost all atmospheric environmental monitoring data is used to prepare environmental reports such as daily newspapers and annual reports (Arsie et al, 2010;Marino et al, 2017) while the value of data needs to be further explored (Li et al, 2014). For example, through historical monitoring data, the future trend of atmospheric environmental quality can be predicted so as to better guide people's production activities (Chien et al, 2005(Chien et al, , 2010. Meanwhile, it provides relevant scientific evidence for government decision-making and management departments in the formulation of systems concerning this area (Chien et al, 2003).…”

Section: Introductionmentioning

confidence: 99%

Analysis Technology of Environmental Monitoring Data Based on Internet of Things Environment and Improved Neural Network Algorithm

Zhai¹

2019

Appl. Ecol. Env. Res.

View full text Add to dashboard Cite

With the development of industrialization, the problem of environmental pollution has become increasingly serious. Environmental monitoring data, as a measurement index of environmental quality, is increasingly valued by governments and citizens. However, the environmental data accumulated through real-time monitoring at the current stage is mostly used to write basic reports, while the hidden laws or values still need to be further explored. This paper proposes two original environmental prediction models and improves upon these two methods separately to predict the quality of the atmospheric environment. The following research results are obtained: the multivariate linear equation is optimized through stepwise linear regression, which can accurately predict the short-term atmospheric environmental quality; the improved BP neural network can predict the mid-term and long-term atmospheric environmental quality through short-term training.

show abstract

A Feasibility Study of a Predictive Emissions Monitoring System Applied to Taipower’s Nanpu and Hsinta Power Plants

Cited by 8 publications

References 11 publications

Development of Predictive Emissions Monitoring System Using Open Source Machine Learning Library – Keras: A Case Study on a Cogeneration Unit

Development of Predictive Emissions Monitoring System Using Open Source Machine Learning Library – Keras: A Case Study on a Cogeneration Unit

Long-term evaluation of machine learning based methods for air emission monitoring

Analysis Technology of Environmental Monitoring Data Based on Internet of Things Environment and Improved Neural Network Algorithm

Contact Info

Product

Resources

About