Design And Implementation Of Three-Dimensional Localization System Based On Uncertainty Analysis

Wireless localization is one of the key technologies of wireless sensor networks,and it has been widely applied in many fields such as military,industry,transportation,and people's livelihood.With the advent of artificial intelligence,internet of things,and intelligent driving times,wireless localization technology has immeasurable application prospects.Compared with 2D localization,3D localization can provide more accurate and richer location information,but there are still technical challenges in terms of positioning accuracy,stability,and availability,including non-line-of-sight error in communication distance estimation,localization uncertainty,localization visualization and other key problems.Research on 3D localization technology has important theoretical significance and application value.Based on a thorough review of current researches on 3D localization technology at home and abroad,this dissertation focus on designing and developing a 3D localization system.A nanoTrack-V2.0 development kit manufactured by Nanotron Company of Germany is applied as the hardware platform for the 3D localization system,and the localization software is developed based on LabWindows/CVI independently.The 3D localization system can achieve accurate real-time positioning and image display in 3D scenes.Firstly,we study the 3D localization algorithm,and propose a feature point-based estimated distance filtration method for the problem of non-line-of-sight error in communication distance estimation.This method works on the 3D localization operations,through a voting mechanism to achieve the screening and elimination of the estimated distance with large non-line-of-sight error,so as to weaken the negative impact of non-line-of-sight propagation of wireless signals on the positioning results and improve the positioning accuracy.The results of simulation and experiments show that the feature point-based estimated distance filtration method can improve the average positioning accuracy by 26.96%.Then,we analyze the uncertainty in 3D localization process.The GUM(Guide to expression of Uncertainty in Measurement)method and the MCM(Monte Carlo Method)method are used to evaluate the uncertainty components in the three coordinate dimensions,and the uncertainty is intergrated with the form of 2-norm.The simulation and experimental analysis results show that the relative error of the uncertainty obtained from GUM and MCM is kept within 5%,which verify the accuracy and feasibility of the two methods.Finally,based on the above theoretical research,a 3D localization software is developed by using LabWindows/CVI tools.This software is the control,calculation,and display center of the 3D localization system.Accurate positioning results can be obtained in the test under complex environment,and the positioning results and their uncertain information are displayed in the visual interface accurately and visually.The innovations and contributions to the research field of this dissertation mainly include the following three points:(1)The proposed feature point-based estimated distance filtration method can improve the positioning accuracy in complex environments significantly,which provide new ideas for solving the non-line-of-sight error of distance estimation in the field of wireless localization.(2)The concept of uncertainty is introduced into the 3D localization field.The uncertainty information indicates the trustworthiness of positioning results,which provides important prior knowledge for subsequent decisions and navigation.(3)The developed localization software has an image-based 3D visualization interface,which has certain guidance and reference value for the design and development of other localization software.