Research On Map Completion And Autonomous Positioning Method Based On Inertial Navigation

With the rapid popularity of smartphones,high-precision pedestrian autonomous location services are rapidly developing and people can already enjoy location services using their smartphones on-the-go.Generally speaking,Wi Fi,Bluetooth and zigbeebased pedestrian navigation technologies require the advance acquisition of accurate indoor flat maps to assist and build specific databases.However,in practical applications,it is often encountered that the environment is unknown or an accurate and complete indoor map is not available.How to quickly build indoor maps at low cost in the absence of indoor positioning deployment prerequisites is one of the research hotspots in this field.In this thesis,an inertial guidance-based indoor map building method is investigated for partial missing map scenarios,which can quickly complete indoor map completion in combination with pedestrian inertial guidance under the condition that only partial maps are known.The research contents of the thesis are as follows.(1)First,for the special scenario where only part of the map is known,a mapcompletion algorithm is studied,which can extract effective information from the incomplete map and combine it with pedestrian inertial guidance trajectories to achieve map completion.(2)Secondly,a particle filtering algorithm is used to partially match the pedestrian inertial guidance trajectory with the known part of the map,a weighted rotation angle algorithm is studied to realize the initial direction correction of the original inertial guidance trajectory,and the map corridor width is estimated by using a large number of scattered particles.(3)Finally,a map anchor point search method is studied based on pixel point detection and particle filtering algorithm,which determines two anchor points of the map and then combines the acquired map parameters with the pedestrian inertial trajectory to construct an accurate indoor map,effectively reducing the location and size errors that exist when the inertial trajectory is used to construct an indoor map.The algorithm is able to locate pedestrians in the unmapped part of the map and complete the missing part of the constructed map at the same time when only part of the map is known.By comparing with the real environment indoor map,the indoor map completion method based on indoor pedestrian inertial navigation data and incomplete maps in this thesis,the mapped corridor is closer to the original map corridor.The three average indexes under the comparison of the mapped and real maps are obtained by testing and verifying 192 experimental samples: the map overlap matching degree is92.13%,the proportion of the extended part is 4.23%,and the proportion of the reduced part is 3.23%.The percentage of the reduced part is 3.64%.The indoor map can be built quickly and inexpensively with guaranteed accuracy.