Acta Optica Sinica, Volume. 38, Issue 12, 1229001(2018)
Method for Direct Temperature Extraction of Brillouin Scattering Spectra Based on Radial Basis Function Neural Network
Fig. 1. Brillouin scattering gain spectra
Fig. 2. Flow chart of temperature measurements based on RBFNN and fitting methods
Fig. 3. Structural diagram of RBFNN
Fig. 4. Structural diagram of Brillouin gain spectrum test system
Fig. 5. Brillouin spectra at different sweeping frequencies and at 45 ℃. (a) 1 MHz; (b) 5 MHz; (c) 10 MHz; (d) 20 MHz
Fig. 6. Smooth fitting curve of Brillouin spectrum at 45 ℃
Fig. 7. Linear fitting of Brillouin frequency shift
Fig. 8. Errors of test samples using fitting method
Fig. 9. Test errors of RBFNN and BPNN at sweeping frequency of 0.175 MHz
Fig. 10. Predicted temperature deviations at 5 different sweeping frequencies
Fig. 11. RMSEs for 3 methods at 5 different sweeping frequencies
Fig. 12. RMSEs for 3 methods after linewidth extension at 5 different sweeping frequencies
|
Get Citation
Copy Citation Text
Yang Sui, Chuannan Meng, Wei Dong, Xindong Zhang. Method for Direct Temperature Extraction of Brillouin Scattering Spectra Based on Radial Basis Function Neural Network[J]. Acta Optica Sinica, 2018, 38(12): 1229001
Category: Scattering
Received: Mar. 30, 2018
Accepted: Jul. 28, 2018
Published Online: May. 10, 2019
The Author Email: