Indoor And Outdoor Seamless Positioning Method Based On Multi-sensor Fusion

Positioning technology is the key technology in intelligent vehicle,mobile robot and other fields.At present,the outdoor positioning mainly uses the Global Navigation Satellite System(GNSS).However,when the positioning target is severely shielded,especially in the indoor environment,the positioning error will increase rapidly.The above mobile devices always need to travel back and forth in the indoor and outdoor environments many times during the working process.Therefore,the indoor and outdoor seamless positioning technology has received great attention at home and abroad.This thesis conducts research on indoor/outdoor seamless positioning and indoor positioning,and uses multi-sensor fusion as a method to solve the positioning data uncertainty problem caused by the independence of indoor and outdoor positioning technologies.In order to avoid installing extra equipment on positioning target and improve the adaptability of the positioning method,this thesis proposes an indoor positioning method based on target tracking.In this method,the target tracking equipment is deployed indoor environment,the positioning target is detected by sensors,and the target identity and motion state are determined by data association and tracking,and then the target positioning information is sent to the corresponding mobile terminal to realize indoor positioning.Different from the typical target tracking methods,in order to maintain accurate target identity information in a multi-target scenario,this method needs to determine the identity correspondence tables of multiple target trajectories and multiple mobile terminals at the initial time of tracking.Besides,when the target moves in multiple indoor positioning environments,in order to maintain the target trajectory across the region,a target trajectory maintenance method with the trajectory and measurement shared is proposed to realize the continuous and stable tracking and positioning in the large range of indoor environments.In order to solve the problem of indoor and outdoor seamless positioning,this thesis designs a multi-sensor fusion indoor and outdoor seamless positioning method.This method uses federated filters to fuse multiple positioning data.Inertial Navigation System(INS)was taken as the public system.Outdoor positioning sub-filter was constructed by INS and GNSS,and indoor positioning sub-filters were constructed by INS and several indoor positioning data respectively.The above sub-fulters are used to construct a global federated filter to achieve indoor and outdoor seamless positioning.In order to solve the adaptability problem of positioning methods in different environments,an adaptive federated filter is proposed in this thesis.This filter can adaptively adjust the information distribution parameters and measurement noise of the sub-filter according to the sensor data,so as to reduce the influence of the positioning result with large error data.Finally,this thesis designs and develops the indoor and outdoor seamless positioning system.The experiment results show that the system can achieve higher positioning accuracy in indoor and outdoor environment,especially at the boundary between indoor and outdoor,which verifies the effectiveness of the system and the method.