Laser & Optoelectronics Progress, Volume. 56, Issue 13, 130601(2019)
Feature Extraction and Recognition Algorithm for Fiber Intrusion Signals
Figures & Tables(10)
Fig. 1. Processing flow chart of feature extraction and recognition for fiber intrusion signals
Fig. 4. Preprocessing for original intrusion signals. (a1)(b1) Tapping signal; (a2)(b2) vehicle signal; (a3)(b3) running signal
Fig. 5. Results of energy ratio. (a) Tapping signal; (b) vehicle signal; (c) running signal
Fig. 6. Variance analysis results. (a) Tapping signal; (b) vehicle signal; (c) running signal
Table 1. Model error analysis under different λ, w, and b
View tableTable 1. Model error analysis under different λ, w, and b
λ w ,b ∈[-2, 2] w ,b ∈[-100, 100] w ,b ∈[-200, 200] w ,b ∈[-400, 400] 0.005 0.278 0.165 0.198 0.323 0.050 0.292 0.176 0.202 0.210 0.500 5.000 0.297 0.312 0.162 0.186 0.172 0.189 0.176 0.203
Table 2. Comparison of recognition results
View tableTable 2. Comparison of recognition results
Recognition method Recognition result /% RVFL neural network BP neural network 96.7 94.7
Tools
Get Citation
Copy Citation Text
Hongquan Qu, Dianjun Gong, Changnian Zhang, Yanping Wang. Feature Extraction and Recognition Algorithm for Fiber Intrusion Signals[J]. Laser & Optoelectronics Progress, 2019, 56(13): 130601
Paper Information
Category: Fiber Optics and Optical Communications
Received: Dec. 12, 2018
Accepted: Jan. 24, 2019
Published Online: Jul. 11, 2019
The Author Email: Gong Dianjun (769353964@qq.com)
© Copyright 2018-2021 | Chinese Laser Press.
All Rights Reserved 沪ICP备15018463号-20