Application of probabilistic deep learning models to simulate thermal power plant processes

dc.contributor.advisorLaubscher, Ryno
dc.contributor.authorRaidoo, Renita Anand
dc.date.accessioned2023-04-20T11:13:20Z
dc.date.available2023-04-20T11:13:20Z
dc.date.issued2022
dc.date.updated2023-04-18T09:33:03Z
dc.description.abstractDeep learning has gained traction in thermal engineering due to its applications to process simulations, the deeper insights it can provide and its abilities to circumvent the shortcomings of classic thermodynamic simulation approaches by capturing complex inter-dependencies. This works sets out to apply probabilistic deep learning to power plant operations using historic plant data. The first study presented, entails the development of a steady-state mixture density network (MDN) capable of predicting effective heat transfer coefficients (HTC) for the various heat exchanger components inside a utility scale boiler. Selected directly controllable input features, including the excess air ratio, steam temperatures, flow rates and pressures are used to predict the HTCs. In the second case study, an encoder-decoder mixturedensity network (MDN) is developed using recurrent neural networks (RNN) for the prediction of utility-scale air-cooled condenser (ACC) backpressure. The effects of ambient conditions and plant operating parameters, such as extraction flow rate, on ACC performance is investigated. In both case studies, hyperparameter searches are done to determine the best performing architectures for these models. Comparisons are drawn between the MDN model versus standard model architecture in both case studies. The HTC predictor model achieved 90% accuracy which equates to an average error of 4.89 W m2K across all heat exchangers. The resultant time-series ACC model achieved an average error of 3.14 kPa, which translate into a model accuracy of 82%.
dc.identifier.apacitationRaidoo, R. A. (2022). <i>Application of probabilistic deep learning models to simulate thermal power plant processes</i>. (). ,Faculty of Engineering and the Built Environment ,Department of Mechanical Engineering. Retrieved from http://hdl.handle.net/11427/37790en_ZA
dc.identifier.chicagocitationRaidoo, Renita Anand. <i>"Application of probabilistic deep learning models to simulate thermal power plant processes."</i> ., ,Faculty of Engineering and the Built Environment ,Department of Mechanical Engineering, 2022. http://hdl.handle.net/11427/37790en_ZA
dc.identifier.citationRaidoo, R.A. 2022. Application of probabilistic deep learning models to simulate thermal power plant processes. . ,Faculty of Engineering and the Built Environment ,Department of Mechanical Engineering. http://hdl.handle.net/11427/37790en_ZA
dc.identifier.ris TY - Master Thesis AU - Raidoo, Renita Anand AB - Deep learning has gained traction in thermal engineering due to its applications to process simulations, the deeper insights it can provide and its abilities to circumvent the shortcomings of classic thermodynamic simulation approaches by capturing complex inter-dependencies. This works sets out to apply probabilistic deep learning to power plant operations using historic plant data. The first study presented, entails the development of a steady-state mixture density network (MDN) capable of predicting effective heat transfer coefficients (HTC) for the various heat exchanger components inside a utility scale boiler. Selected directly controllable input features, including the excess air ratio, steam temperatures, flow rates and pressures are used to predict the HTCs. In the second case study, an encoder-decoder mixturedensity network (MDN) is developed using recurrent neural networks (RNN) for the prediction of utility-scale air-cooled condenser (ACC) backpressure. The effects of ambient conditions and plant operating parameters, such as extraction flow rate, on ACC performance is investigated. In both case studies, hyperparameter searches are done to determine the best performing architectures for these models. Comparisons are drawn between the MDN model versus standard model architecture in both case studies. The HTC predictor model achieved 90% accuracy which equates to an average error of 4.89 W m2K across all heat exchangers. The resultant time-series ACC model achieved an average error of 3.14 kPa, which translate into a model accuracy of 82%. DA - 2022_ DB - OpenUCT DP - University of Cape Town KW - Air-cooled condensers KW - Natural convection boilers Time-series prediction KW - Deep learning KW - Mixture density networks KW - Recurrent neural networks LK - https://open.uct.ac.za PY - 2022 T1 - Application of probabilistic deep learning models to simulate thermal power plant processes TI - Application of probabilistic deep learning models to simulate thermal power plant processes UR - http://hdl.handle.net/11427/37790 ER - en_ZA
dc.identifier.urihttp://hdl.handle.net/11427/37790
dc.identifier.vancouvercitationRaidoo RA. Application of probabilistic deep learning models to simulate thermal power plant processes. []. ,Faculty of Engineering and the Built Environment ,Department of Mechanical Engineering, 2022 [cited yyyy month dd]. Available from: http://hdl.handle.net/11427/37790en_ZA
dc.language.rfc3066eng
dc.publisher.departmentDepartment of Mechanical Engineering
dc.publisher.facultyFaculty of Engineering and the Built Environment
dc.subjectAir-cooled condensers
dc.subjectNatural convection boilers Time-series prediction
dc.subjectDeep learning
dc.subjectMixture density networks
dc.subjectRecurrent neural networks
dc.titleApplication of probabilistic deep learning models to simulate thermal power plant processes
dc.typeMaster Thesis
dc.type.qualificationlevelMasters
dc.type.qualificationlevelMSc
Files
Original bundle
Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
thesis_ebe_2022_raidoo renita anand.pdf
Size:
13.35 MB
Format:
Adobe Portable Document Format
Description:
License bundle
Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
license.txt
Size:
0 B
Format:
Item-specific license agreed upon to submission
Description:
Collections