Research On Indoor Positioning Algorithm Based On PDR And Geomagnetic Matching Fusion

Since the 21 st century,the technology of global navigation satellite system(GNSS)has developed rapidly and tends to be mature.In the gorgeous city,the transportation network brings great convenience to people's life,and the reliable location-based service technology becomes more and more important.Due to the increase of indoor environment such as high-rise buildings,tunnels and mines,indoor positioning technology has been greatly applied.The complex indoor environment will seriously restrict the satellite signal,resulting in unsatisfactory GNSS positioning results,and even unreliable positioning problems.Therefore,domestic and foreign researchers hope to use various methods to establish a reliable indoor positioning system.This paper presents an indoor pedestrian positioning system based on geomagnetic field,pressure field,gravity field and other geophysical fields.According to the movement characteristics of pedestrians,the system extracts the movement characteristics,judges the movement state of pedestrians,and then realizes pedestrian dead reckoning(PDR).However,the inertial navigation system has the problems of high precision in a short time and long-term drift.The system introduces geomagnetic matching technology,uses particle filter algorithm for fusion positioning,and finally obtains better pedestrian position information.In terms of height positioning,this paper relies on atmospheric pressure and gravity field to achieve floor positioning and stair positioning.The main research contents and innovations of this paper are listed as follows:1.In the PDR algorithm,by analyzing the walking characteristics of pedestrians ?,this paper proposes to use the constraint form of time amplitude to detect the step frequency,identify the peak and valley of pedestrian acceleration in each step,and obtain the time between the maximum acceleration amplitude and the peak and valley in this step.By setting the threshold form to constrain these two values,we can finally obtain more accurate pedestrian steps.Considering that the pedestrian step size estimation is an important factor affecting the PDR algorithm,this paper improves the nonlinear step size model,adds time variables to the famous Weinberg model,and uses the form of double model coefficients to further improve the step size estimation accuracy.2.This paper introduces the construction method of geomagnetic fingerprint map.In the divided geomagnetic acquisition area,a number of paths with the same direction are set,and pedestrians collect and record every step of geomagnetic data at a constant speed.Compared with the traditional grid acquisition method,this method is simple and efficient.Finally,the fusion algorithm based on particle filter is described,which makes the PDR and geomagnetic matching complement each other to obtain more accurate pedestrian trajectory.3.In this paper,location-based service is extended to three-dimensional space,which can make use of the characteristics of air pressure change and overweight and weightlessness for floor positioning.When the pedestrian moves in the stairs,the state of going up and down stairs is identified by analyzing the change characteristics of acceleration and air pressure.Then the PDR algorithm is extended to three-dimensional space according to the angle and height of stairs.Finally,the accurate three-dimensional position information of pedestrian is obtained by combining with geomagnetic matching algorithm.4.While studying and learning the indoor fusion positioning technology,we cooperate to develop hardware devices suitable for pedestrians.In the integrated development environment of Matlab guide,the graphical interface of fusion positioning is designed.In practical application,the software has strong reliability and practicability,which brings great convenience to the experiment.