Modeling of Relative Exergy Destruction for Turboprop Engine Components Using Deep Learning Artificial Neural Networks

Tolga Baklacioglu; Onder Turan; Hakan Aydin

doi:10.1515/tjj-2018-0047

Article

Modeling of Relative Exergy Destruction for Turboprop Engine Components Using Deep Learning Artificial Neural Networks

Tolga Baklacioglu , Onder Turan and Hakan Aydin

Published/Copyright: January 19, 2019

Published by

Become an author with De Gruyter Brill

Author Information

From the journal International Journal of Turbo & Jet-Engines Volume 38 Issue 4

Abstract

This study illustrates an deep learning approach supported by a metaheuristic design targeting the foremost features and parameters of artificial neural network (ANN) framework used in predicting relative exergy destruction ( ( f r e l , d e s t * ) ) of a turboprop components. The development of deep ANN comprising of three-hidden layers using data obtained considering multiple engine input parameters was accomplished. Once the deep learning ANN frameworks were hybridized with a metaheuristic approach, such as genetic algorithms (GAs). The analysis of errors revealed a close fit involving the predicted values of the model and references made on data in f r e l , d e s t ∗ for the engine main components. The use of appropriately chosen values in preceding networks weights produced more accurate testing results (Linear correlation coefficient values (R) for the engine components are found to be between 0.996497 and 0.998986) in networks using three hidden layers compared to those using lower hidden layers. Furthermore, optimizing deep ANNs using GAs delivers not only further improved accuracy (R is calculated to be in the range of 0.998929–0.999966 for the engine components) but also an effective utilization of time in the resulting models.

Keywords: turboprop; energy; relative exergy destruction; artificial neural networks; deep learning; genetic algorithms

PACS: Entropy thermodynamics; 05.70.-a; Air transportation; 89.40.Dd; Irreversible thermodynamics; 05.70.Ln; Energy conservation in classical mechanics; 45.20.dh; Thermal analysis; 81.70.Pg; Thermodynamic considerations; 88.05.De

GLOSSARY

Abbreviations
ANN: Artificial neural network
BP: Back-propagation
Comb: Combustor
CV: Cross-validation
GA: Genetic algorithm
gen: Generated
HL: Hidden layer
MAE: Mean absolute error
MLP: Multilayer perceptron
MSE: Mean squared error
nMSE: Normalized mean squared error
per: Perfect
T: Training
Symbols
a: Air
b: Bias constant
C: Compressor
ch: Chemical
Cp: Specific heat
dest: Destruction
d_k: k_th component of the desired output vector
e: Specific energy
e_k: Error function of neuron k
ex: Specific exergy rate
E: Energy rate; Mean squared error function of output layer
Ex: Exergy rate
f: Fuel
gg: Gas generator turbine
h_n, y_j: Output of the n^th (j^th) neuron in the hidden layer
h_pr: Fuel heating value
i: initial
in: Inlet
k: k^th component; Number of nodes in the output layer
K: Number of output processing elements
N: Number of heat reservoirs
o_k: k^th component of the actual output vector
out: Outlet
P: Pressure; Pattern size for relative exergy destruction
ph: Physical
pt: Power turbine
R: Specific gas constant; Linear correlation coefficient
rev: Reversible
s: Specific entropy
S: Entropy
t: Iteration number
tot: Total
u: Useful
v: Specific volume
V: Volume
v_ji: Weight value in the connection between the jth hidden and the ith output neurons
x_i: Input layer
w_n,i: Weight between the i^th input neuron and n^th hidden neuron
w_m,n: Weight between the n^th hidden neuron and m^th neuron in the output layer
Δw: Weight increment
W: Work transfer interaction
Y_actual: Actual value of the relative exergy destruction
Y_m: Output of the m^th neuron in the output layer
Y_mean: Mean value of the relative exergy destruction
Y_predicted: Predicted relative exergy destruction
β: Momentum factor
η: Step-size (learning rate)
φ: Transfer function
φ_j’(α_j): Derivative of the transfer function of hidden layer
φ_k’(α_k): Derivative of the transfer function of output layer
λ: Combustion equation constant
0: Reference (environment) state

References

1. Topal A, Turan O. On-Design exergy analysis of a combustor for a turbojet engine: a parametric study. Aircr Eng Aerosp Tec. 2017;89:719–24.10.1108/AEAT-12-2016-0254Search in Google Scholar

2. Balli O. Advanced exergy analyses of an aircraft turboprop engine (TPE). Energy. 2017;124:599–612.10.1016/j.energy.2017.02.121Search in Google Scholar

3. Cheze B, Gastineau P, Chevallier J. Forecasting world and regional aviation jet fuel demands to the mid-term (2025). Energy Policy. 2011;39:5147–58.10.1016/j.enpol.2011.05.049Search in Google Scholar

4. Sahin O, Turan O. Evaluation of aircraft descent profile. Energy Procedia. 2016;95:308–13.10.1016/j.egypro.2016.09.011Search in Google Scholar

5. Meric OS. The proposal of the arrival route for the metroplex airports. Journal of Aeronautics and Space Technologies. 2015;8:19–26.Search in Google Scholar

6. Meric OS. Optimum arrival routes for flight efficiency. Journal of Power and Energy Engineering. 2015;3:449–52.10.4236/jpee.2015.34061Search in Google Scholar

7. Aydin H, Turan O. Numerical calculation of energy and exergy flows of a turboshaft engine for power generation and helicopter applications. Energy. 2016;115:914–23.10.1016/j.energy.2016.09.070Search in Google Scholar

8. Cilgin ME, Turan O. Entropy generation calculation of a turbofan engine: a case of CFM56-7B. Int J Turbo Jet Eng. 2017;58:664–71.10.1515/tjj-2017-0053Search in Google Scholar

9. Atashkari K, Nariman-Zadeh N, Pilechi A, Jamali A, Yao X. Thermodynamic pareto optimization of turbojet engines using multi-objective genetic algorithms. Int J Therm Sci. 2005;44:1061–71.10.1016/j.ijthermalsci.2005.03.016Search in Google Scholar

10. Homaifar A, Lai HY, McCormic E. System optimization of turbofan engines using genetic algorithms. Appl Math Model. 1994;18:72–83.10.1016/0307-904X(94)90162-7Search in Google Scholar

11. Liu F, Sirignano WA. Turbojet and turbofan engine performance increases through turbine burners. J Propul Power. 2001;17:695–705.10.2514/6.2000-741Search in Google Scholar

12. Balli O. Afterburning effect on the energetic and exergetic performance of an experimental turbojet engine (TJE). Int J Exergy. 2014;14:205–36.10.1504/IJEX.2014.060278Search in Google Scholar

13. Balli O, Aras H, Aras N, Hepbasli A. Exergetic and exergoeconomic analysis of an aircraft jet engine (AJE). Int J Exergy. 2008;5:567–81.10.1504/IJEX.2008.020826Search in Google Scholar

14. Turan O. Exergetic effects of some design parameters on the small turbojet engine for unmanned air vehicle applications. Energy. 2012;46:51–61.10.1016/j.energy.2012.03.030Search in Google Scholar

15. Turan O. Effect of reference altitudes for a turbofan engine with the aid of specific-exergy based method. Int J Exergy. 2012;11:252–70.10.1504/IJEX.2012.049738Search in Google Scholar

16. Turan O. An exergy way to quantify sustainability metrics for a high bypass turbofan engine. Energy. 2015;86:722–36.10.1016/j.energy.2015.04.026Search in Google Scholar

17. Balli O, Hepbasli A. Energetic and exergetic analyses of T56 turboprop engine. Energy Convers Manag. 2013;73:106–20.10.1016/j.enconman.2013.04.014Search in Google Scholar

18. Baklacioglu T, Aydin H, Turan O. Energetic and exergetic efficiency modeling of a cargo aircraft by a topology improving neuro-evolution algorithm. Energy. 2016;103:630–45.10.1016/j.energy.2016.03.018Search in Google Scholar

19. Baklacioglu T, Turan O, Aydin H. Dynamic modeling of exergy efficiency of turboprop engine components using hybrid genetic algorithm-artificial neural networks. Energy. 2015;86:709–21.10.1016/j.energy.2015.04.025Search in Google Scholar

20. Kotas TJ. The exergy method of thermal plant analysis, Reprint ed. Florida, USA: Krieger, Malabar, 1995.Search in Google Scholar

21. Bejan A, Tsatsaronis G, Moran M. Thermal design and optimization. USA: Wiley, 1996.Search in Google Scholar

22. Svorcan J, Stupar S, Trivkovic S, Petrašinovic N, Ivanov T. Active boundary layer control in linear cascades using CFD and artificial neural networks. Aerosp Sci Technol. 2014;39:243–9.10.1016/j.ast.2014.09.010Search in Google Scholar

23. Hu Q, Zhang R, Zhou Y. Transfer learning for short-term wind speed prediction with deep neural networks. Renew Energ. 2016;85:83–95.10.1016/j.renene.2015.06.034Search in Google Scholar

24. Li X, Yang Y, Pang Z, Wu X. A comparative study on selecting acoustic modeling units in deep neural networks based large vocabulary Chinese speech recognition. Neurocomputing. 2015;170:251–6.10.1007/978-3-642-42057-3_60Search in Google Scholar

25. Hinton G, Deng L, Yu D, Dahl G, Mohamed A, Jaitly N, et al. Deep neural networks for acoustic modeling in speech recognition. IEEE Signal Proc Mag. 2012;29:82–97.10.1109/MSP.2012.2205597Search in Google Scholar

26. Bourlard H, Morgan N. Connectionist speech recognition: a hybrid approach. Norwell, MA: Kluwer, 1993.10.1007/978-1-4615-3210-1Search in Google Scholar

27. Glorot X, Bengio Y Understanding the difficulty of training deep feed-forward neural networks. In: Proc. AISTATS, 2010: 249–56.Search in Google Scholar

28. Goldberg DE. Genetic algorithms in search, optimization and machine learning. Reading, MA: Addison-Wesley, 1989.Search in Google Scholar

29. Jin Y, Olhofer M, Sendhoff B. A framework for evolutionary optimization with approximate fitness functions. IEEE T Evolut Comput. 2002;481–94.10.1109/TEVC.2002.800884Search in Google Scholar

30. Wang L. A hybrid genetic algorithm–neural network strategy for simulation optimization. Appl Math Comput. 2005;170:1329–43.10.1016/j.amc.2005.01.024Search in Google Scholar

31. Adewole BZ, Abidakun OA, Asere AA. Artificial neural network prediction of exhaust emissions and flame temperature in LPG (liquefied petroleum gas) fueled low swirl burner. Energy. 2013;61:606–11.10.1016/j.energy.2013.08.027Search in Google Scholar

32. Hosoz M, Ertunc HM. Artificial neural network analysis of an automobile air conditioning system. Energ Convers Manage. 2006;47:1574–87.10.1016/j.enconman.2005.08.008Search in Google Scholar

33. Asadi E, Gameiro da Silva M, Antunes CH, Dias L, Glicksman L. Multi-objective optimization for building retrofit: A model using genetic algorithm and artificial neural network and an application. Energ Buildings. 2014;81:444–56.10.1016/j.enbuild.2014.06.009Search in Google Scholar

34. Togun NK, Baysec S. Prediction of torque and specific fuel consumption of a gasoline engine by using artificial neural networks. Appl Energ. 2010;87:349–55.10.1016/j.apenergy.2009.08.016Search in Google Scholar

35. Malleswaran M, Vaidehi V, Sivasankari N. A novel approach to the integration of GPS and INS using recurrent neural networks with evolutionary optimization techniques. Aerosp Sci Technol. 2014;32:169–79.10.1016/j.ast.2013.09.011Search in Google Scholar

Received: 2018-12-16

Accepted: 2019-01-09

Published Online: 2019-01-19

Published in Print: 2021-12-20

You are currently not able to access this content.

Articles in the same Issue

https://doi.org/10.1515/tjj-2018-0047

Keywords for this article

turboprop; energy; relative exergy destruction; artificial neural networks; deep learning; genetic algorithms