| Visible Light Positioning(VLP)is a new generation of Positioning technology integrating intelligent lighting and optical communication.Compared with other traditional indoor Positioning technologies,VLP has been widely concerned for its advantages of strong electromagnetic interference resistance,high spectrum utilization,and green environmental protection.The technology based on the received signal strength RSS(Received Signal Strength)algorithm is the most widely used technology in the visible light indoor positioning system.It can ensure a relatively good positioning effect at a lower cost.Indoor reflection interference,and at the same time,there is the limitation that the superposition of each light source signal cannot accurately separate the light source information,which makes it difficult to achieve accurate positioning in complex indoor environments.Therefore,based on RSS positioning,this paper focuses on the light source layout,spectral estimation detection multi-light source signal separation algorithm,and neural network positioning model construction.(1)The channel model of the visible light indoor positioning system is built.The link modes of LOS(Line Of Sight)and NLOS(Not Line Of Sight)are studied respectively.Considering that only one reflection in NLOS has a great influence on the system,this paper only studies the channel gain of the direct link and the first reflection path.The light source layout provides a theoretical basis.(2)The optimal light source layout of the visible light positioning system is obtained by simulation comparison.Three light source layout models of 3LED array,4LED array,and5 LED array are built,and each model is simulated to analyze and compare the minimum illuminance,illuminance standard deviation and illuminance distribution data of each scheme.Finally,based on international lighting standards,comprehensive consideration Light source layout illumination,uniformity and positioning effect,choose the 4LED array model as the light source layout method of this system.(3)On the basis of RSS positioning,an indoor positioning method based on the combination of Power Spectrum Detection(PSD)and Back Propagation Neural Networks(BPNN)is proposed.The spectral estimation detection algorithm can solve the problem of difficult separation of multi-light source signals,and obtain the frequency information loaded by each LED through the separation information;the neural network can fit the functional relationship between optical power and distance,and then combine the error constraint function to predict the final position coordinates.(4)Build a simulation and measurement platform to verify the algorithm proposed in this paper theoretically and practically.On the basis of theory and simulation positioning,a visible light real-time positioning platform is built to verify the algorithm theory and simulation.The results show that: in the multi-light source signal separation stage using spectrum estimation detection,the maximum error of the four groups of source frequency detection is 1.253 Hz,and the maximum power error is 0.125 W;in the simulation positioning stage,a 4m × 4m × 3m three-dimensional space is built,respectively at H = 0 m,H=0.5 m,H=1.0 m three plane heights for positioning test,the average positioning error is 3.81cm;in the actual positioning stage,a0.8×0.8×0.8m solid wood stereotaxic model is built,and the average error is 4.28 cm.The research in this paper provides an effective way to separate the signal energy of multiple light sources,overcome the background noise,and improve the indoor visible light positioning accuracy. |
PDF Full Text Request