| Recently,with the development of the internet of things industry,location-based services are increasingly becoming an important part of people’s social life.According to statistics,people spend most of their time indoors,and the application of indoor positioning technology is becoming more and more important.At present,many indoor positioning technologies are used.Due to its higher bandwidth,lower power consumption,insensitivity to channel fading,strong anti-interference,and strong penetrability,the positioning accuracy of Ultra Wide Band(UWB)can reach 0.1-0.5m,which is an indoor positioning technology with relatively high positioning accuracy.However,the positioning accuracy of UWB is affected by clock synchronization accuracy,non-line-of-sight factors,human absorption,and obstructions,and the positioning results will have large errors.As an economical and simple positioning method,indoor map data can not only visually display the location of pedestrians on the map,but also have a natural binding force on the positioning results without any other auxiliary facilities or instruments.Therefore,it has important research significance and value on how to realize the efficient integration of UWB positioning results and map information.In this topic,the indoor positioning algorithm based on map matching is studied.Based on the map information matrix,the corresponding through-wall detection algorithm and cross-region detection algorithm are proposed.Then based on these two detection algorithms,a map matching algorithm based on particle filter multiple weights update is proposed.The details are as follows:Firstly,the current research status of indoor positioning technology and map matching algorithm at home and abroad is introduced.The deficiencies of existing algorithms are analyzed,and the main research content of this topic is summarized.Secondly,the basic knowledge of UWB positioning and map matching algorithm is introduced.The positioning principle of UWB is described,three existing map matching algorithms are analyzed,and the classic particle filter algorithm theory is explained in detail.Thirdly,the detection algorithm based on the map information matrix is researched.In view of the large amount of calculation of the map matching algorithm based on particle filter,the map modeling process is improved,the map information matrix is obtained by using the rasterized map,and the idea of crowdsourcing is introduced to improve the construction efficiency of the map information matrix.Then a wall-through detection algorithm based on the map information matrix is proposed,which reduces the amount of calculation and shortens the running time of the algorithm on the basis of ensuring that the wall-through phenomenon is correctly detected.Aiming at the error cross-region phenomenon in the particle update process of the map matching algorithm based on particle filter,a cross-region detection algorithm based on the map information matrix is proposed.The accuracy and effectiveness of the through-wall detection algorithm and the cross-area detection algorithm are verified by simulation and measured data.Finally,the assisted localization algorithm based on map matching is researched.Aiming at the outliers caused by the loss of UWB positioning signal or other factors,a weighted first-order extrapolation smoothing algorithm is proposed to preprocess the outliers of UWB on the basis of analyzing various smoothing algorithms.For the problem that the positioning result is located in an unreachable area,based on the through-wall detection algorithm and the cross-region detection algorithm based on the map information matrix,the weight of the location is added,and a map matching algorithm based on particle filter multi-weight update is proposed.The traditional weights are improved to be composed of distance weights,area weights,wall penetration weights,and cross-area weights.Theoretical analysis and actual measurement results verify the rationality and effectiveness of the improved algorithm. |
PDF Full Text Request