BES-RBF神经网络支持下的可见光定位系统

Research of visible light positioning system based on BES-RBF neural network

随着科技的飞速发展,在理想可见光信道模型中进行定位已无法满足实际应用需求。本文建立的定位系统模型考虑了可见光的直射链路和不规则墙壁的一次反射,并基于该模型提出了一种秃鹰搜索算法优化径向基神经网络(BES-RBF)的单LED灯的定位算法。该定位系统首先通过秃鹰搜索算法优化径向基神经网络的初始权值和阈值,使得优化后的RBF神经网络结构更加稳定;然后将接收器各PD(Power Delivery)接收到的光功率值构成的指纹库数据,引入优化后的RBF网络模型中进行训练,建立定位模型;最后获得定位信息,再通过WKNN对部分点进行优化,获得精确的位置信息。仿真试验表明,在3 m×3 m×3 m的空间内,本文算法的平均定位误差为5.54 cm,80%的定位误差在4.5 cm内,与其他定位算法相比,定位精度有了显著提升。

With the rapid development of science and technology,positioning in the ideal visible channel model can no longer meet the practical application requirements.The positioning system model established in this paper considers the direct link of visible light and a reflection of irregular walls.Based on this single LED lamp model,a bald eagle search algorithm is proposed to optimize of radial basis network(BES-RBF).The positioning system first optimizes the initial weights and thresholds of the radial basis function(RBF) neural network through the bald eagle search algorithm to make the structure of the optimized RBF neural network more stable.Then,the fingerprint database data consisting of the light power values received by each PDs of the receiver is introduced into the optimized RBF network model for training,and the positioning model is established.After the location information is obtained,some points are corrected by WKNN to obtain the final location information.In 3 m×3 m×3 m space,the simulation results show that,the average positioning error of the algorithm proposed in this paper is 5.54 cm,and 80% of the positioning error is within 4.5 cm.Compared with other positioning algorithms,the positioning accuracy has been significantly improved.