Positioning correction method based on Bluetooth signal single-step robust filtering and approximation discrimination

Abstract

The invention relates to the field of indoor positioning, and discloses a positioning correction method based on Bluetooth signal single-step robust filtering and approximation discrimination. The method comprises the steps of: building a Bluetooth node network; carrying out parameter initialization configuration; performing gait recognition according to the accelerometer data; dividing and analyzing Bluetooth signal data according to a single-step time period to obtain a Bluetooth node single-step signal intensity sequence; performing single-step signal intensity sequence robust processing onthe signal intensity sequence; respectively comparing the robust single-step signal intensity values with a signal intensity identification threshold value; sorting the single-step signal intensity values of observable Bluetooth nodes, and carrying out approximation judgment on the maximum signal intensity value; and carrying out proximity correction. According to the method, robust processing iscarried out on the signal intensity sequence of the observable Bluetooth node obtained in each step through a single-step robust filtering method, the gross error influence in the signal intensity value is reduced, large-amplitude jitter is eliminated, the problems of frequent error correction and track transfer are avoided, and the pedestrian walking positioning correction precision and reliability are improved.

Description

technical field [0001] The invention relates to the technical field of indoor positioning, in particular to a positioning correction method based on Bluetooth signal single-step robustness filtering and approach discrimination. Background technique [0002] At present, many WLAN devices are deployed in indoor environments. Although no additional deployment is required and the signal coverage is relatively large, there is a strong multipath effect. WiFi positioning usually uses fingerprint positioning technology, but fingerprint-based positioning requires manpower and time to collect a large amount of training data in the offline stage to build a fingerprint database. Changes in indoor environment AP or indoor layout will affect the quality of the fingerprint database. In recent years, with the widespread popularity of smart phones, more and more sensors are integrated into mobile phones, and mobile phones can obtain more data in the environment, which further promotes the de...

AI Technical Summary

Problems solved by technology

In 2016, Wang Gechao and others proposed the acceleration difference (AD-FSM) algorithm to improve the method of the finite state machine, which can accurately count steps under the interference state; in 2015, Kang et al. combined the gyroscope and the magnetic field sensor and proposed the Smart PDR; but the above-mentioned purely relying on mobile phone sensor data for positioning will cause serious cumulative error problems. In 2015, Yang Fan of Shanghai Jiaotong University and others used the positioning method of PDR and WiFi results. PDR data can be used for fingerprint positioning and fingerprint database matching. scope

Due to the fluctuation range of the transmission power of the Bluetooth beacon, and the signal is easily blocked by pedestrians during the signal propagation process, the Bluetooth signal strength observed by the mobile phone sensor fluctuates.

The bouncing bluetooth observation signal strength brings the uncertainty of bluetooth proximity recognition, which leads to large deviations, jumps, errors, etc. in the PDR calibration results

[0004] However, the classic threshold discrimination method faces the difficulty of selecting the discrimination threshold in Bluetooth proximity discrimination.

When the threshold value is set too high, the correction range is reduced, and the possibility of users entering the correction area is reduced. There is a problem that the correction point is not recognized and the correction is missed.

When the threshold is set too low, the correction range is enlarged, and there are problems such as premature correction of the user position, large correction errors, obvious jumps of the user trajectory during correction, and stagnation of the user trajectory within a large correction range.

Images