Research On Localization Algorithm Of Underground Parking Lot Based On RFID

The rapid growth of the economy,the increasing number of cars on the road has resulted in a shortage of parking spaces,particularly in urban areas with large commercial and residential buildings.To meet the demands of modern life,more and more underground parking lots are being constructed.However,due to the complex structure of these parking lots,many car owners face difficulties in parking and finding their cars,which affects the efficiency and experience of using underground parking spaces.As a result of these parking problems,the efficiency and experience of underground parking have been seriously affected.The GPS,the Beidou and other location technologies can no longer meet people’s requirements for the location accuracy of underground parking lots.Therefore,to address these challenges,a high-precision indoor location method is urgently needed for vehicle location.This has led to an increased research focus on indoor location technology,with RFID in the Internet of Things being a popular trend due to its affordability and accuracy.This thesis proposes three algorithms to improve the accuracy of RFID location systems in underground parking lots.1)Aiming at the static location problem in the grid deployment scenario of reference tags,an TLOW algorithm is proposed.The TLOW algorithm enhances the accuracy of the trilateral location algorithm by introducing the observation source RFID reader.The algorithm gives a credibility weight to the rough location result based on the distance from the observation source to the tag being tested,improving location accuracy by 26.18% and system stability by 38.45% compared to the trilateral location algorithm.2)Aiming at the static location problem in the scenario of sparse and irregular deployment of reference tags,an LAVT algorithm is proposed.The LAVT algorithm is based on the LANDMARC algorithm.The LAVT algorithm inserts virtual tags at the circumcenter to gradually reduce the area of the nearest neighbor region of the tag being tested.The algorithm also proposes to use the second-order Lagrange interpolation method to estimate the RSSI of virtual tags more accurately.In cases of sparse and irregular deployment of reference tags,the location accuracy is improved by 19.03% compared to LANDMARC.3)Aiming at the dynamic location problem in the moving scene of the tag to be located,an MTTA algorithm is proposed.The algorithm employs an optimal layout method combining reference tags and antenna arrays,and the Kalman filter is used to filter tag signals in the measured area,creating a fingerprint database of reference tags.The algorithm measures the angle of the moving vehicle through the antenna associated with the dynamic neighbor tag,and then measures the trajectory of the moving vehicle.The location accuracy of the MTTA algorithm for moving vehicles is increased by 68.08% and 42.26% compared to the EKF＿WSN algorithm and the WSL algorithm.Overall,the three algorithms proposed in this thesis demonstrate their effectiveness in enhancing the accuracy of RFID location systems and are well-suited for underground parking lots.