Research On Major Technologies Of Indoor Positioning Based On RFID,Inertial Sensors And Magnetic Features

Location information plays an increasingly important role in the fields of intelligent perception,security supervision,logistics and so on.The real-time,accurate and continuous location information acquisition has become a prerequisite and critical issue for location-based services(LBS),especially for the management of people and objects in semi-enclosed places,such as mines,large-scale event venues,schools,construction sites,etc.However,due to the influence of environmental factors,obstacle obscuration and thermal noise of sensors,there are still some key problems to be solved in practical applications of indoor localization.(1)The fingerprint based indoor localization is becoming a dominant solution for its high applicability in complex indoor environment.However,the extensive site survey efforts on manpower and time have become a major bottleneck.(2)In indoor positioning based on wireless technology,when the object moves in a line of sight(LOS)and non-line of sight(NLOS)mixed environment,the fluctuation of RSS will cause big positioning bias.How to effectively identify LOS/NLOS environment and reduce positioning error is a key issue to improve wireless-based positioning accuracy.(3)In a complex indoor environment,how to effectively combine advantages of wireless technology and MEMS sensors to achieve an accurate,stable positioning solution has always been a challenge in the indoor positioning field.Based on research on the key issues in indoor positioning,this thesis proposes an indoor positioning with three different technologies: radio frequency identification(RFID),inertial motion sensor and magnetic sensor.The solution achieves an average positioning accuracy of 1.96 meters.The major contributions of this dissertation are as follows:1)It presents a zero-effort fingerprint automated construction and site survey update scheme.With the crowdsourcing method,it puts forward a method to automatically build and update the fingerprint database based on RFID,pedestrian dead reckoning(PDR)and magnetic matching(MM).In order to solve the problem that step length would vary from person to person which results in positioning bias,the RSS technology and floor map is introduced to aid step length estimation.Experiments show that after 200 mins of algorithm execution,the obtained fingerprint database is proven to achieve the similarity accuracy as the manually labeled database.The similarity of the RSS fingerprint database is 75%,and the similarity of magnetic field characteristics is 90%.2)It presents a method for identification of the LOS / NLOS environment and reduction of NLOS propagation error.It mainly improves on three aspects: inertial measurement unit(IMU)assisted received signal strength(RSS)filtering,IMU-assisted LOS/NLOS environment identification and NLOS error reduction.The calculation of Cramer-Rao lower bound(CRLB)theoretically proves that the positioning accuracy obtained by proposed method is significantly better than positioning accuracy obtained only by using RSS information.Experiments show that the reliability of two different environment switching identifications is 95% on average,and it can effectively reduce the positioning error caused by NLOS.3)It proposes an indoor positioning method based on the fusion of RFID and multi-sensor technologies.Fully utilizing the complementary features of RFID,PDR and MM technologies,the gradual refinement of the coarse-grained to fine-grained positioning method is realized.In this work,the different localization regions in the indoor environment are abstracted into the corresponding corridor model or subarea model.The corresponding positioning models in different positioning areas are used to achieve accurate indoor positioning.Experiments show that the proposed method has an average error of 1.96 meters,which is 175% and 82.6% higher than the method of RFID fingerprint positioning and magnetic field matching respectively.