Research On Target Positioning And Tracking Based On UAV And RFID Technology

With the development of UAV technology,the use of UAVs for target search has been widely used in military strategy,industrial manufacturing and smart scenic spots,and has very broad application prospects.However,traditional UAV target positioning and tracking usually relies on computer vision technology,which cannot accurately locate the target in the case of non-line-of-sight,which greatly affects the efficiency of positioning.This paper studies the target positioning and tracking system based on quadrotor UAV and Radio Frequency Identification(RFID)technology,and uses the UAV carrying the RFID reader to identify the target with the electronic tag for localization and tracking.In this paper,algorithms such as Received Signal Strength Indicator(RSSI)preprocessing,positioning and tracking of targets in space,and path planning of UAVs are studied,which not only give full play to the flexibility of UAVs,but also avoid the constraints of non-line-of-sight environments.It will play an important role in application scenarios such as ruin search and rescue,prison monitoring,and guardianship of the elderly and children.The main research contents of this paper are as follows:(1)First,the composition and principle of the RFID positioning system are expounded,and the common positioning algorithms based on ranging and non-ranging are analyzed,and RSSI is selected as the basic data for positioning from the perspective of improving positioning accuracy and reducing system cost.Then,according to the hardware circuit,the reader software is designed to read the RSSI data in the RFID electronic tag,and then the S500 UAV is assembled,and finally the target positioning and tracking hardware platform based on UAV and RFID technology is constructed.(2)To improve the positioning accuracy,based on the logarithmic path loss model and the attenuation factor model,the analysis and research are carried out from three aspects: data preprocessing,RF signal propagation model optimization and positioning algorithm.First,in the preprocessing stage,the performance of the commonly used single filter is evaluated and the advantages of each filter algorithm are integrated,and an improved hybrid filter method is designed and implemented.Secondly,in the optimization processing stage of the model,the correlation coefficient in the path loss equation is optimized by using the two electronic tags arranged in advance as a reference.Finally,in terms of positioning algorithm,for the case of drones locating indoor targets from outdoors,by comparing the commonly used three-dimensional space coordinate calculation methods,the least squares method of Taylor series processing with constraints is proposed to solve the position coordinates,and the MATLAB simulation software and experiments in the actual environment verify the feasibility of the algorithm.(3)Aiming at the trajectory planning problem of UAV,firstly establishing the dynamic model of UAV and using cascade PID method to control the trajectory of UAV.For stationary targets,it is mainly divided into coarse-grained positioning process and fine-grained positioning process.The trajectory planning in the coarse-grained positioning process is mainly based on static path planning based on four non-coplanar points to complete the solution of the coordinates.In the process of fine-grained positioning,the greedy algorithm is mainly used to select the point of the maximum RSSI value on the circumference as the initial point,and then the distance converted from RSSI is always used as the step size and the potential field method is used to approach the target object.Until it flies to the vicinity of the target,the search for the target is completed.(4)For the study of moving objects,firstly,the models of uniform motion,uniform acceleration motion and variable acceleration motion are established.Then,according to the changeable motion state of the target in space,the commonly used tracking filtering algorithm is analyzed,and the IMM-PF tracking algorithm is proposed based on the interactive multi-model combined with the particle filtering algorithm.In addition,to deal with the problem of easy particle degradation in the process of particle filtering,an improved genetic algorithm is used to improve the diversity of particles in the process of genetic algorithm.Finally,the UAV completes the tracking of the moving target in an iterative circle manner and MATLAB software is used to complete the verification of the algorithm.