Research On Geomagnetic Indoor Positioning Algorithm Based On Particle Filter

As the demand for location-based services becomes more apparent,the importance of indoor positioning becomes more prominent.Existing indoor positioning solutions include Wi-Fi,Bluetooth,Ultra-wideband,geomagnetic,and so on.Among them,the geomagnetic positioning technology based on fingerprint matching which does not require additional equipment investment and signal stability is receiving more and more attention.Particle filter algorithm is a commonly used matching algorithm in geomagnetic indoor positioning system.However,when using geomagnetic intensity for geomagnetic feature matching,the positioning result error will be too large,and even the positioning loss phenomenon will occur.Aiming at this problem,this paper proposes a matching mode based on geomagnetic sequence.The geomagnetic feature uses geomagnetic intensity,and the Particle Filter is improved by Dynamic Time Warp algorithm and Correlation Coefficient.In the experimental process,the positioning error of particle filter algorithm is firstly analyzed by the measurement robot loaded with HMC5983 three-axis magnetic sensor.Based on MATLAB the huge error of the particle filter algorithm in the case of using only the geomagnetic intensity and the loss of positioning are also shown in detail.At the same time,based on the same data,the improved algorithm is used to compare and analyze the experiment.The result shows that the positioning error is less than 1 m,which solves the problem of positioning loss.Moreover,smart phones integrate a variety of sensors including geomagnetic sensors.In order to verify the scalability of the algorithm,this paper intends to use the smart phone for actual positioning.Before using the smart phone for geomagnetic indoor positioning,the sensor data acquisition app was developed based on the Nokia7 smartphone using the Android platform.The built-in sensors used are geomagnetic sensors,acceleration sensors and game vector sensors.Then the user holds the Nokia7 mobile phone and positions in the corridor.Finally,the experiment shows that the improved particle filter algorithm based on smart phone has better accuracy than 3 meters after positioning convergence.The phenomenon of localization distortion in particle filtering is improved in the improved algorithm,and the improved algorithm has practical generalization.