Anti-NLOS interference visible light positioning method and system based on linear regression model

Abstract

The invention discloses an anti-NLOS interference visible light positioning method and system based on a linear regression model. The anti-NLOS interference visible light positioning system comprisesa transmitting module, a receiving module, a training module and a positioning module. The transmitting module comprises a plurality of LED lamps, a signal generator, a baseband modulation module anda driving circuit; each LED lamp is provided with an ID; the receiving module comprises an optical filter, a photoelectric detector, an amplifier, an analog-to-digital converter and a microprocessor,and is used for receiving signals of different LED lamps in different time slices in one period and extracting the intensity of the received signals. According to the method, non-line-of-sight transmission (NLOS) interference is removed by adopting a machine learning method, and then, positioning is realized by adopting a geometric positioning method, so that the influence of the NLOS interferenceon positioning can be eliminated, the positioning precision is improved, and the requirement of an existing fingerprint positioning method based on machine learning on positioning data is also reduced.

Description

technical field [0001] The invention belongs to the technical field of positioning, and in particular relates to a linear regression model-based anti-NLOS (non-line-of-sight transmission) interference visible light positioning method and system thereof. Background technique [0002] With the wide application of positioning and navigation technology and the continuous diversity of location service applications, indoor positioning technology has gradually entered people's field of vision, and the related demand has also continued to grow. Existing indoor positioning technologies mainly include: infrared technology, ultrasonic technology, Bluetooth technology, WiFi technology, ultra-wideband technology, etc. But these methods do not have a positioning system that can simultaneously have the advantages of low cost, high positioning accuracy, simple structure, and low power consumption, and does not have the characteristics of wide application. [0003] In recent years, with the...

AI Technical Summary

Problems solved by technology

[0004] However, for VLC positioning in indoor environments, the geometric positioning method is low in cost and simple to implement, and has been widely used, but it is seriously affected by NLOS (Not line of sight, non-line of sight transmission) interference, resulting in a decrease in positioning accuracy, especially on walls. Surrounding locations are particularly affected by NLOS interference, which often leads to a sharp drop in positioning accuracy

Aiming at the above-mentioned interference problem, the VLC fingerprint positioning method based on machine learning has been widely adopted, such as "A high-precision indoor visible light positioning method based on artificial neural network (application number CN106100734B)", this type of method directly uses multiple received signal strengths to Estimating the position has achieved good results, but it needs to collect enough data for training, and the training effect will directly affect the positioning accuracy

Images