Fig. 1. Extreme learning machine algorithm single hidden layer feedforward neural network

Fig. 2. Schematic diagram of the machine learning and prediction procedures

Fig. 3. Temperature and concentration Gaussian distribution model
(a): Temperature distribution; (b): Concentration distribtion

Fig. 4. Flow field fan light layout

Fig. 5. Test set error

Fig. 6. Test set determination factor

Fig. 7. Integrated absorbance prediction error of harmonic signal S_2f/1f added to noise

Fig. 8. Direct-connected scramjet flow field parallel light layout

Fig. 9. Schematic diagram of the measurement system of the direct-connected scramjet
(a): Direct-connected superesonic combustion test bed; (b): TDLAS measurement system

Fig. 10. Frequency division multiplexing wavelength modulated absorption signal

Fig. 11. Time sequence of Experiment

Fig. 12. Experimental integral absorbance prediction results

Fig. 13. Flow field temperature variation with time