Mass-pedestrian Seamless Positioning Technology With Fusing Smartphone And Foot-mounted PDR

The pedestrian navigation system(PNS)is a new application positioning technology that emerged in the late 1990 s.It uses a dedicated device to realize real-time positioning and tracking of individuals and guide pedestrians and to find your destination quickly.Global Navigation Satellite System(GNSS)provides users with continuous,highprecision position,speed and time information in a globally open environment.In recent years,with the rapid development of new Micro Electro Mechanical System(MEMS)technology and its increasing popularity in mobile intelligent terminals,we can also use the Pedestrian Dead Reckoning(PDR)method to obtain continuous relative pose information.PNS combines satellite navigation and positioning technology with PDR based on MEMS sensors to achieve complementary advantages and provide high-precision,continuously available pedestrian navigation services for mass users.This paper systematically studies the seamless pedestrian positioning technology of smart phone and foot-mounted PDR.A sensor fusion algorithm of low-cost GNSS and MEMS is proposed for the problem that GNSS signals may be attenuated or even interrupted due to weak signal or non-line of sight in urban complex scenes.The quality of observations of low-cost GNSS equipment is analyzed in detail,and the corresponding data preprocessing and filtering positioning methods are proposed.The zero-speed correction is combined with the step-length-heading(SL-Heading)PDR.Under the constraint of multi-source observation,high-precision dead reckoning is realized to provide continuous relative pose.Based on the built-in multi-source sensor of Android smartphone,the pedestrian navigation application software "Walker" was developed.The main works and contributions of the thesis are as follows:(1)Evaluated the data quality characteristics of Android GNSS observations,and detailed the methods of obtaining observations,such as satellite visibility,carrier-tonoise ratio,pseudo-range noise,pseudo-range rate,phase rate,and Doppler.Using precision phase and Doppler observations to smooth pseudorange is expected to achieve high precision positioning.(2)A GNSS filtering positioning algorithm for joint motion model is proposed.Based on the above data quality analysis,a more robust GNSS observation data preprocessing strategy is presented.Gross error detection and culling are performed by using the consistency of the time differenced observation,and unified into the pseudorange rate/phase rate/Doppler velocity measurement equation.The state model of position change is established by using the speed measurement information,and the pseudorange standard point positioning(SPP)is used to measurement update.The field test results show that the static positioning error RMS of the filter positioning algorithm in the ENU direction is 0.60 m,0.54 m and 1.36 m,respectively.Compared with the results of the smartphone GNSS chipset,the position accuracy increased by 47% and 20% in the horizontal and up-vertical directions,respectively.(3)The application of MEMS sensor observations in PDR is discussed in detail.The important steps of SL-Heading PDR and INS-ZUPT PDR methods are summarized.The adaptive zero-speed detection,barometer height constraint and magnetic heading constraint are deeply studied,and continuous 3D-PDR based on multi-source observations is realized,which is useful for indoor and outdoor seamless pedestrians.The field test results show that the position error of the 3D-PDR in the horizontal and up-vertical directions is 1.65 m and 0.46 m,the cumulative distance error accuracy is 0.97%,and the up-vertical closure difference reaches the decimeter level accuracy.(4)The indoor and outdoor seamless positioning algorithm based on Android GNSS/MEMS fusion is proposed.The GNSS speed is used to improve the PDR positioning performance and the mathematical model of GNSS/PDR fusion filter positioning is given.The field test shows that the fusion filtering algorithm has the ability to seamlessly locate indoor and outdoor in complex scenes of pedestrian navigation.The cumulative position error of the TDCP is 0.51%,which can effectively correct the PDR error parameters.(5)Designed and implemented the “Walker” Android application software,based on the Android operating system development,embedded self-developed GNSS/PDR fusion positioning algorithm,which can realize seamless navigation and positioning of indoor and outdoor users.The field test results show that the horizontal position error of the fusion positioning in the partially occluded scene(the playground and the main road)is better than 2 m,which is double the accuracy of the output of the smartphone GNSS chip;in the completely occluded scene(the teaching building tunnels and indoor)reached the decimeter-level up-vertical closure accuracy,and the cumulative distance error accuracy is better than 1%;the smartphone GNSS and the foot-mounted MEMS fusion achieved 2 m-level indoor and outdoor seamless positioning and smooth switching.