HTFF_178.pdf (593.16 kB)
A neuro-fuzzy model of evaporator in organic rankine cycle
conference contribution
posted on 2023-06-10, 03:47 authored by Hamid Enayatollahi, Peter FusseyPeter Fussey, Bao Kha NguyenBao Kha NguyenThe Organic Rankine Cycle (ORC) is a propitious waste heat recovery (WHR) technology that allows recovery of wasted energy from low to medium temperature sources. This WHR method needs to be adopted as an Internal Combustion Engine (ICE) bottoming technology to mitigate its environmental effects and fulfil exhaust gas emission regulations. The evaporator is the most decisive element of the ORC cycle due to its high nonlinear behaviour and high thermal inertia. In this study, a neuro-fuzzy model of the evaporator is presented based on the data obtained from Finite Volume (FV) model of the evaporator. The simulation results are compared in terms of RMSE, error mean and standard deviation. The data obtained from ANFIS model reached a promising agreement with FV model. For prediction of the evaporator outlet temperature, RMSEs of 0.152 and 1.33 obtained for the training and test data, respectively. Furthermore, the ANFIS model was successfully able to predict the evaporator power with RMSE of 0.035 for the training and 0.2 for the test data. In addition, the ANFIS model compared to the FV model with twenty control volumes enhanced the simu lation time significantly. This clearly indicates the great potential of employing ANFIS model for real-time applications.
History
Publication status
- Published
File Version
- Published version
Journal
Proceedings of the World Congress on Mechanical, Chemical, and Material EngineeringISSN
2369-8136Publisher
Avestia PublishingExternal DOI
Page range
178Article number
HTFFEvent name
The 5th World Congress on Mechanical, Chemical, and Material EngineeringEvent location
Lisbon, PortugalEvent type
conferenceEvent date
15 - 17 Aug 2019Place of publication
Ottawa, CanadaISBN
9781927877616Department affiliated with
- Engineering and Design Publications
Full text available
- Yes
Peer reviewed?
- Yes
Legacy Posted Date
2022-06-06First Open Access (FOA) Date
2022-06-10First Compliant Deposit (FCD) Date
2022-06-06Usage metrics
Categories
No categories selectedKeywords
Licence
Exports
RefWorks
BibTeX
Ref. manager
Endnote
DataCite
NLM
DC