Aeroengine thrust estimation is an important problem for direct thrust control since it is unmeasurable. Many methods and algorithms have been proposed to solve this problem. Unfortunately, almost all these methods can only estimate aeroengine thrust when the engine is in steady state. Hence, this study proposes an algorithm based on long short-term memory networks and gradient boosting for aeroengine thrust estimation in transition state. The newly proposed algorithm can estimate thrust of an aeroengine when its working state is changed from one steady state to another. The experimental results demonstrated that the proposed algorithm can be well applied to estimate aeroengine thrust in transition state and the estimated precision can meet the requirements of thrust estimation.
ZhaoY-PSunJ-G. Aeroengine thrust estimation using least squares support vector regression machine. J Aerospace Power2009; 24(6): 1420–1425.
2.
GastineauZHappawanaG. Robust model-based control for jet engines. Aiaa/asme/sae/asee Joint Propulsion Conference and Exhibit, 1998.
3.
MaggioreMOrdóñezRPassinoKM , et al.Estimator design in jet engine applications. Eng Appl Artif Intelligence2003; 16(7): 579–593.
4.
YaoY-LSunJ-G. Application of adaptive genetic neural network algorithm in design of thrust estimator. J Aerospace Power2007; 22: 1748–1753.
5.
Krishna KumarKYachisakoYHuangY. Jet engine performance estimation using intelligent system technologies. 39th Aerospace Sci Meet Exhibit2001; 1: 1122.
6.
LittJS. An optimal orthogonal decomposition method for kalman filter-based turbofan engine thrust estimation. J Eng Gas Turbines Power2008; 130(1): 745–756.
7.
HenrikssonMGrönstedtTBreitholtzC. Model-based on-board turbofan thrust estimation. Control Eng Pract2011; 19(6): 602–610.
LiuY-NZhangS-XZhangC. Aero engine thrust estimator design based on kernel method. J Propulsion Tech2013; 34(6): 829–835.
10.
SuW-SZhaoY-PSunJ-G. Novel weighted least squares support vector regression for thrust estimation on performance deterioration of aero-engine. Trans Nanjing Univ Aeronaut Astronaut2012; 29(1): 7.
11.
ZhaoY-PSunJ-GWangQ-G, et al.Thrust estimator design based on K-means clustering and reduced least squares support vector regression. J Aerospace Power2010; 25(5): 1177–1183
12.
EberhartSY-HParticle swarm optimization: developments, applications and resources. In: Congress on Evolutionary Computation, Seoul, South Korea, 27–30 May 2001.
13.
HuangG-BZhuQ-YSiewC-K. Extreme learning machine: Theory and applications. Neurocomputing2006; 70(1–3): 489–501.
14.
SongH-QLiB-WZhangY, et al.Aero-engine thrust estimator design based on clustering and particle swarm optimization extreme learning machine. J Propulsion Tech2017; 38(6): 1379–1385.
15.
BengioYFrasconiPSimardP. The problem of learning long-term dependencies in recurrent networks. In: IEEE International Conference on Neural Networks, San Francisco, CA, 28 March–1 April 1993.
16.
HochreiterSSchmidhuberJ. Long short-term memory. Neural Comput1997; 9(8): 1735–1780.
17.
MayerHGomezFWierstraD, et al.A system for robotic heart surgery that learns to tie knots using recurrent neural networks. Adv Robotics2008; 22(13–14): 1521–1537.
18.
SchmidhuberJWierstraDGomezFJ. Evolino: Hybrid neuroevolution/optimal linear search for sequence prediction. In: Proceedings of the 19th International Joint Conferenceon Artificial Intelligence, Scotland, UK, 30 July–5 August 2005.
19.
FernándezSGravesASchmidhuberJ. An application of recurrent neural networks to discriminative keyword spotting. Int Conf Artif Neural Networks2007; 1: 220–229.
20.
GravesASchmidhuberJ. Framewise phoneme classification with bidirectional LSTM and other neural network architectures. Neural Networks2005; 18(5–6): 602–610.
21.
GersFASchmidhuberJCumminsF. Learning to forget: Continual prediction with LSTM. Neural Comput1999; 22(10): 2451–2471.
RumelhartDEHintonGEWilliamsRJ. Learning representations by back-propagating errors. Nature1986; 323(6088): 533–536.
28.
MartensJSutskeverI. Learning recurrent neural networks with hessian-free optimization. In: Proceedings of the 28th International Conference on Machine Learning (ICML-11), Bellevue, WA, 28 June–2 July 2011.
29.
PascanuRMikolovTBengioY. On the difficulty of training recurrent neural networks. International Conference on Machine Learning2013; 28: 1310–1318.
30.
ArdalanifarsaMZolfaghariS. Chaotic time series prediction with residual analysis method using hybrid Elman-NARX neural networks. Neurocomputing2010; 73(13): 2540–2553.
31.
ChandraRZhangM. Cooperative coevolution of Elman recurrent neural networks for chaotic time series prediction. Neurocomputing2012; 86: 116–123.
32.
CrowderRSIII. Predicting the Mackey-Glass Timeseries with Cascade-Correlation learning. Netherlands: Elsevier, 1991, pp. 117–123.
33.
ShiZ-WHanM. Support vector echo-state machine for chaotic time-series prediction. IEEE Trans Neural Networks2007; 18(2): 359–372.
