A hybrid of fast K-nearest neighbor and improved directed acyclic graph support vector machine for large-scale supersonic inlet flow pattern recognition

Abstract

Inlet flow pattern recognition is one of the most crucial issues and also the foundation of protection control for supersonic air-breathing propulsion systems. This article proposes a hybrid algorithm of fast K-nearest neighbors (F-KNN) and improved directed acyclic graph support vector machine (I-DAGSVM) to solve this issue based on a large amount of experimental data. The basic idea behind the proposed algorithm is combining F-KNN and I-DAGSVM together to reduce the classification error and computational cost when dealing with big data. The proposed algorithm first finds a small set of nearest samples from the training set quickly by F-KNN and then trains a local I-DAGSVM classifier based on these nearest samples. Compared with standard KNN which needs to compare each test sample with the entire training set, F-KNN uses an efficient index-based strategy to quickly find nearest samples, but there also exists misclassification when the number of nearest samples belonging to different classes is the same. To cope with this, I-DAGSVM is adopted, and its tree structure is improved by a measure of class separability to overcome the sequential randomization in classifier generation and to reduce the classification error. In addition, the proposed algorithm compensates for the expensive computational cost of I-DAGSVM because it only needs to train a local classifier based on a small number of samples found by F-KNN instead of all training samples. With all these strategies, the proposed algorithm combines the advantages of both F-KNN and I-DAGSVM and can be applied to the issue of large-scale supersonic inlet flow pattern recognition. The experimental results demonstrate the effectiveness of the proposed algorithm in terms of classification accuracy and test time.

Keywords

Supersonic inlet flow pattern recognition machine learning k-nearest neighbors directed acyclic graph support vector machine big data

Get full access to this article

View all access options for this article.

References

Oates

. Aircraft Propulsion Systems Technology and Design. USA: American Institute of Aeronautics and Astronautics, 1989.

Chen

Tan

H-J

Zhang

Q-F

, et al. Buzz flows in an external-compression inlet with partially isentropic compression. AIAA J 55 2017: 4286–4295.

Oswatitsch

. Pressure recovery for missiles with reaction propulsion at high supersonic speeds (the efficiency of shock diffusers). Contributions to the Development of Gasdynamics. Wiesbaden: Vieweg Teubner Verlag, 1980, pp. 290–323.

Trapier

Duveau

Deck

. Experimental study of supersonic inlet buzz. AIAA J 2006 44(10): 2354–2365.

Dailey

. Supersonic diffuser instability. J Aeronaut Sci 1955 22(11): 733–749.

Lee

Kim

, et al. Flow characteristics of small-sized supersonic inlets. J Propul Power 2011; 27(2): 306–318.

Voland

Auslender

Smart

, et al. CIAM/NASA Mach 6.5 scramjet flight and ground test. In: 9th international space planes and hypersonic systems and technologies conference, Reston, Virigina. Norfolk, VA: AIAA, 1999.AIAA Paper:1999-4848

Khelif

Chebel-Morello

Malinowski

, et al. Direct remaining useful life estimation based on support vector regression. IEEE Trans Ind Electron 2016; 64(3): 2276–2285.

Wang

. An intelligent diagnosis scheme based on generative adversarial learning deep neural networks and its application to planetary gearbox fault pattern recognition. Neurocomputing 2018; 310: 213–222.

10.

Trapier

Deck

Duveau

, et al. Time-frequency analysis and detection of supersonic inlet buzz. AIAA J 2007; 45(9): 2273–2284.

11.

Chang

Bao

Xie

. Optimal classification criterions of hypersonic inlet start/unstart. J Propul Power 2007; 23: 310–316.

12.

Chang

Bao

Fan

. Operation pattern classification of hypersonic inlets. Acta Astronautica 2009; 65: 457–466.

13.

Chang

Bao

, et al. A CFD assessment of classifications for hypersonic inlet start/unstart phenomena. Aeronaut J 2009; 113: 263–271.

14.

Chang

, et al. Classifier utility modeling and analysis of hypersonic inlet start/unstart considering training data costs. Acta Astronautica 2011; 69: 841–847.

15.

Chang

Zheng

, et al. Minimization of classification samples for supercritical and subcritical patterns of supersonic inlet. J Therm Sci 2014; 23: 375–380.

16.

Chang

, et al. Recent research progress on unstart mechanism, detection and control of hypersonic inlet. Prog Aerospace Sci 2017; 89: 1–22.

17.

Zhang

Zong

, et al. Efficient KNN classification with different numbers of nearest neighbors. IEEE Trans Neural networks Learn Syst 2018; 29: 1774–1785.

18.

Liu

C-L

Nakagawa

. Evaluation of prototype learning algorithms for nearest-neighbor classifier in application to handwritten character recognition. Pattern Recognition 2001; 34: 601–615.

19.

Domeniconi

Jing Peng

Gunopulos

. Locally adaptive metric nearest-neighbor classification. IEEE Trans Pattern Anal Machine Intelligence 2002; 24: 1281–1285.

20.

Bhattacharya

Ghosh

Chowdhury

. An affinity-based new local distance function and similarity measure for KNN algorithm. Pattern Recognition Lett 2012; 33: 356–363.

21.

Wang

Neskovic

Cooper

. Neighborhood size selection in the k-nearest-neighbor rule using statistical confidence. Pattern Recognition 2006; 39: 417–423.

22.

Mullick

Datta

Das

. Adaptive learning-based -nearest neighbor classifiers with resilience to class imbalance. IEEE Trans Neural networks Learn Syst 2018; 29: 5713–5725.

23.

Weinberger

Saul

. Distance metric learning for large margin nearest neighbor classification. J Machine Learn Res 2009; 10: 207–244.

24.

Bar-Hillel

Hertz

Shental

, et al. Learning distance functions using equivalence relations. In: Proceedings of the 20th international conference (ICML 2003), DC, USA, 21–24 August, 2003, 11–18.

25.

Yang

Wang

Zuo

. Fast neighborhood component analysis. Neurocomputing 2012; 83: 31–37.

26.

Belka

Fischer

Pohlmann

, et al. LVQ-KNN: composition-based DNA/RNA binning of short nucleotide sequences utilizing a prototype-based k-nearest neighbor approach. Virus Res 2018; 258: 55–63.

27.

J-L

Wang

L-Z

J-J

, et al. Research and application on KNN method based on cluster before classification. In: 2008 International conference on machine learning and cybernetics, Kunming, China, 12–15 July 2008.

28.

Hsu

Lin

. A comparison of methods for multiclass support vector machines. IEEE Trans Neural Networks 2002; 13: 415–425.

29.

Bottou

Cortes

Denker

, et al. Comparison of classifier methods: a case study in handwritten digit recognition. In: Proceedings of the international conference on pattern recognition, Jerusalem, October 1994, 77–82.

30.

Knerr

Personnaz

Dreyfus

. Single-layer learning revisited: a stepwise procedure for building and training a neural network. In: Soulié FF., Hérault

(eds) Neurocomputing. NATO ASI Series (Series F: Computer and Systems Sciences). Berlin, Heidelberg, Germany: Springer, 1990, pp. 41–50.

31.

Kressel

. Pairwise classification and support vector machines. In: Advances in Kernel Methods: Support Vector Learning. Cambridge, UK: MIT Press, 1998, pp. 255–268.

32.

Platt

Cristianini

Shawe-Taylor

. Large margin DAGs for multiclass classification. Adv Neural Inf Process Syst 2000; 12: 547–553.

33.

Crammer

Singer

. On the learnability and design of output codes for multiclass problems. Machine Learn 2002; 47: 201–233.

34.

Bottou

Vapnik

. Local learning algorithms. Neural Comput 1992; 4: 888–900.

35.

Chen

Tan

H-J

Zhang

Q-F

, et al. Throttling process and buzz mechanism of a supersonic inlet at overspeed mode. AIAA J 2018; 56: 1953–1964.

36.

Chen

Tan

H-J

Liu

Y-Z

, et al. External-compression supersonic inlet free from violent buzz. AIAA J 2019; 57: 2513–2523.

37.

Platt

. Using analytic QP and sparseness to speed training of support vector machines. Adv Neural Inf Process Syst 1999; 11: 557–563.