Fiber-Optic Perimeter Vibration Signal Recognition Based on Local Mean Decomposition and Serial Feature Fusion

Xinglong Xiong; Wantong Zhang; Meng Li; Yuzhao Ma; shuai Feng

doi:10.3788/AOS201939.0206002

Acta Optica Sinica, Volume. 39, Issue 2, 0206002(2019)

Fiber-Optic Perimeter Vibration Signal Recognition Based on Local Mean Decomposition and Serial Feature Fusion

Xinglong Xiong^1、*, Wantong Zhang¹, Meng Li², Yuzhao Ma¹, and shuai Feng³

Author Affiliations

¹ Tianjin Key Laboratory for Advanced Signal Processing, Civil Aviation University of China, Tianjin 300300, China

² Institute of Operation Programming and Safety Technology of Air Traffic Management, Civil Aviation University of China, Tianjin 300300, China

³ Engineering Technical Training Center, Civil Aviation University of China, Tianjin 300300, China

show less

Abstract Get PDF(in Chinese)

Figures & Tables(13)

Fig. 1. Fiber-optic perimeter system based on Mach-Zehnder interferometry

Download full size

Fig. 2. Processing of LMD-ICA

Download full size

Fig. 3. Structural diagram of PNN network

Download full size

Fig. 4. Four typical waveforms of vibration signal. (a) Vibration signal of climbing; (b) vibration signal of knocking; (c) vibration signal of car; (d) vibration signal of natural wind

Download full size

Fig. 5. Comparison of the results of additive reconstruction and ICA reconstruction. (a) Original signal; (b) ICA reconstructed signal; (c) additive reconstructed signal

Download full size

Fig. 6. Feature distributions of different vibration signals. (a) Distribution of K; (b) distribution of Z; (c) distribution of H; (d) distribution of Eapen

Download full size

Fig. 7. Waveforms of different vibration signals in interference environment. (a) Vibration signal in rain; (b) climbing signal in rain; (c) knocking signal in rain

Download full size

Table 1. Effect comparison of the two reconstruction methods
View table
Table 1. Effect comparison of the two reconstruction methods
Reconstruction method Error energy /mV² Signal-to-noise ratio /dB Mean-square error /mV
ICA reconstruction 1.17×10⁵ 11.67 1.71
Additive reconstruction 8.06×10⁵ 3.30 4.49

Table 2. Feature list of four kinds of signals
View table
Table 2. Feature list of four kinds of signals
Classification K Z H E_apen
Climbing 1.69 94 1.98 0.0106
Knocking 1.80 258 2.59 0.1243
Car 1.21 115 2.33 0.0645
Wind 1.33 29 1.72 0.0078

Table 3. Recognition results based on different features

View table

Table 3. Recognition results based on different features

Features	Climbing recognition rate /%	Knocking recognition rate /%	Car recognition rate /%	Wind recognition rate/%	Average recognition rate /%	Average recognition time /s
K,Z	64	80	58	44	61.5	0.76
K,Z,H	92	96	84	74	86.5	0.76
K,Z,H,E_apen	100	100	94	90	96	0.87

Table 4. Recognition results of two methods

View table

Table 4. Recognition results of two methods

Method	Climbing recognition rate /%	Knocking recognition rate /%	Car recognition rate /%	Wind recognition rate /%	Average recognition rate /%	Average recognition time /s
LMD-ICA	100	100	94	90	96.0	0.87
Direct method	90	92	84	80	86.5	0.58

Table 5. Feature list of vibration signals in rain
View table
Table 5. Feature list of vibration signals in rain
Classification K Z H E_apen
Signal in Rain 1.34 6 0.69 0.0025
Climbing signal in rain 1.51 164 1.77 0.0088
Knocking signal in rain 1.69 207 1.85 0.0144

Table 6. Recognition results of two methods in rain

View table

Table 6. Recognition results of two methods in rain

Method	Rain recognition rate /%	Climbing in rain recognition rate /%	Knocking in rain recognition rate /%	Average recognition rate /%	Average recognition time /s
LMD-ICA	100	94	96	96.7	0.91
Direct method	100	84	80	88.0	0.56

Tools

Get Citation

Copy Citation Text

Xinglong Xiong, Wantong Zhang, Meng Li, Yuzhao Ma, shuai Feng. Fiber-Optic Perimeter Vibration Signal Recognition Based on Local Mean Decomposition and Serial Feature Fusion[J]. Acta Optica Sinica, 2019, 39(2): 0206002

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites