Acta Optica Sinica, Volume. 43, Issue 2, 0206004(2023)
Indoor Visible Light Positioning Method Using ISSA-ELM Neural Network Based on Circle Chaotic Mapping
Figures & Tables(14)
Fig. 1. Direct link transmission model of indoor visible light communication channel
Fig. 6. Relationship between RMSE and the number of neurons in the hidden layer of ELM neural network
Fig. 7. Distribution of test points and predicted results for different height receiving planes. (a) h=0 m; (b) h=0.5 m; (c) h=1.0 m; (d) h=1.5 m
Fig. 8. Distribution of positioning errors for different height receiving planes. (a) h=0 m; (b) h=0.5 m; (c) h=1.0 m; (d) h=1.5 m
Fig. 9. Distribution of test points and prediction results of eight positioning methods when the receiver height is 1.5 m. (a) Trilateral measurement method; (b) positioning method using BP neural network; (c) positioning method using BP neural network based on L-M; (d) positioning method using GA-BP neural network; (e) positioning method using SSA-BP neural network; (f) positioning method using ELM neural network; (g) positioning method using SSA-ELM neural network; (h) positioning method using ISSA-ELM neural network
Fig. 10. Cumulative distribution of positioning errors of different positioning methods when the receiver height is 1.5 m
Table 1. Simulation parameters of indoor visible light positioning system
View tableTable 1. Simulation parameters of indoor visible light positioning system
Parameter Value Room size（L×W×H）/（m×m×m） 5×5×3 Position of LEDi（x，y，z）/m LED1：（1，1，3）
LED2：（1，4，3）
LED3：（4，1，3）
LED4：（4，4，3）
Power of each LED bulb /W 4 FOV of LED /（°） 90 Half-power angles of LED Φ1/2 /（°） 60 Effective area of PD A /cm2 1 Gain of optical filter Ts（ψ） 1 Height of receiver /m 0，0.5，1.0，1.5 Data transmission rate R /（Gbit·s-1） 1 Ceiling reflectance ρceil 0.8 Wall reflectance ρwall 0.8 Floor reflectance ρfloor 0.3 Sparrow population number 30 Early warning value ST 0.6 Proportion of discoverers 0.2 Maximum iterations 1000
Table 2. Maximum, minimum, and average positioning errors for different height receiving planes
View tableTable 2. Maximum, minimum, and average positioning errors for different height receiving planes
Receiving plane height h /m 0 0.5 1.0 1.5 Maximum positioning error /cm 3.50 3.60 5.31 12.79 Minimum positioning error /cm 8.8654×10-3 5.9007×10-2 9.9912×10-2 0.3600 Average positioning error /cm 1.01 1.14 1.36 3.87
Table 3. Average positioning errors of eight positioning methods at different height receiving planes
View tableTable 3. Average positioning errors of eight positioning methods at different height receiving planes
Positioning method Average positioning error /cm h=0 h=0.5 m h=1.0 m h=1.5 m Trilateration 124.24 123.18 125.14 132.38 BP 22.84 26.83 44.66 74.66 BP（L-M） 13.10 14.11 15.46 18.03 GA-BP 1.66 1.93 3.12 7.25 SSA-BP 1.45 1.63 2.64 7.09 ELM 1.27 1.42 2.38 6.71 SSA-ELM 1.20 1.33 2.10 4.59 ISSA-ELM 1.01 1.14 1.36 3.87
Table 4. Training time and average positioning time of seven positioning methods based on neural network
View tableTable 4. Training time and average positioning time of seven positioning methods based on neural network
Positioning method Training time /s Average positioning time /s BP 67.0620 0.4356 BP（L-M） 52.6223 0.2416 GA-BP 38.6951 0.0592 SSA-BP 23.5812 0.0157 ELM 0.0700 0.0071 SSA-ELM 0.0480 0.0038 ISSA-ELM 0.0454 0.0035
Tools
Get Citation
Copy Citation Text
Xia Zhao, Junyi Zhang, Qianqian Long. Indoor Visible Light Positioning Method Using ISSA-ELM Neural Network Based on Circle Chaotic Mapping[J]. Acta Optica Sinica, 2023, 43(2): 0206004
Paper Information
Category: Fiber Optics and Optical Communications
Received: Jun. 6, 2022
Accepted: Jul. 21, 2022
Published Online: Feb. 7, 2023
The Author Email: Zhang Junyi (zhangjy@bupt.edu.cn)
© Copyright 2018-2021 | Chinese Laser Press.
All Rights Reserved 沪ICP备15018463号-20