Research On Indoor Positioning Technology Based On Smart Phone

Compared with mature outdoor positioning technology,indoor positioning needs to be improved in terms of implementation cost,positioning complexity,and positioning accuracy.The popularity of smart phones and the variety of sensors built into smart phones have laid the foundation for indoor positioning using smart phones.Taking into account the characteristics of smart phones and the positioning methods suitable for smart phones,this article focuses on indoor geomagnetism positioning alone and the indoor positioning method based on fusion of Wi Fi,PDR(Pedestrian Dead Reckoning)and geomagnetism.The existing indoor geomagnetic positioning technology has many problems to be solved,among which two typical problems are that the large amount of geomagnetic fingerprint data required to be matched is large and the positioning error caused by geomagnetism fingerprint is not unique.This paper proposes two solutions for these two problems: an indoor geomagnetic positioning method based on FCM clustering and location area switching,and an indoor positioning method based on fusion of Wi Fi,PDR and geomagnetism.The main innovations of this article are as follows:This paper introduces the fuzzy clustering algorithm into geomagnetic positioning to reduce the matching amount of geomagnetic fingerprints,thus reducing the calculation amount and accelerating the positioning speed.At the same time,this paper proposes a method of location area switching according to the distribution characteristics of indoor geomagnetic field to reduce the geographic scope of geomagnetic matching to reduce the probability of non-unique geomagnetic fingerprints,and to reduce the amount of geomagnetic fingerprint matching required for each positioning.These two methods can largely solve the two problems mentioned above.After analyzing the three shortcomings of location area switching,this paper proposes the positioning method based on fusion of Wi Fi,PDR and geomagnetism.In the fusion method,through the analysis of the characteristics of the three positioning methods,a new combination of these positioning methods is proposed,which can give full play to the advantages of the three positioning methods.In addition,the maximum likelihood estimation algorithm used for Wi Fi fingerprint matching is weighted and improved,and the positioning accuracy is improved slightly by the weighted maximum likelihood estimation method compared to the ordinary maximum likelihood estimation algorithm.This paper improves the original behavior model according to the walking characteristics of pedestrians.When detecting a sudden turn of a pedestrian,the predicted azimuth in the UKF algorithm is modified by the amount of change in the azimuth measured by the PDR,which Improves filter tracking performance when the azimuth angle changes abruptly,to promote the improvement of positioning accuracy.According to the size of posteriori error of Wi Fi and PDR fusion positioning results,a method of dynamically adjusting the geomagnetic matching range is proposed.The method is based on the standard deviation of the posteriori error of the Wi Fi and PDR fusion positioning results and the current Wi Fi positioning result to jointly adjust the size of the current geomagnetic matching range so as to reduce the scope of the geomagnetic matching under the premise of including the true position of the pedestrian as far as possible.