Research On Multi-source Data Fusion Algorithm And Its Application In Indoor Positioning

The rising economy and progress of science and technology continue to promote thedevelopment of indoor positioning systems. The more complex indoor environments are, dueto the limitations of human vision, the greater demand for tracking mobile terminals orterminal holders in the indoor place. In some areas of real-time tracking, in order to achievehigher accuracy, richer measuring signal sources are often introduced. Combining a variety oftechniques to build positioning system can effectively reduce the shortage of any alone kindof the technologies in improving positioning accuracy, as well as expanding the scope ofpositioning. Faced with a large number of sample data, how to effectively manage andreasonably refined it, is a problem worth exploring. Therefore, multi-source data fusiontechnology is becoming one of the most important indoor positioning technology researches.Based on multi-source data fusion algorithm and its application in indoor positioningoccasions, a popular term is the use of indoor positioning system that can be utilized withmany different positioning technologies. This paper is building sets to form multi-sourcesensor network systems, and explore how to implement a large number of complementaryfusions of information from multiple sensor devices through the optimization algorithm, andultimately improve the overall positioning system performance.Firstly, this paper studies the system structure and measuring principles of ultrasound,infrared, ZigBee signal strength and ultra-wideband, and other positioning technologies aswell as their advantages and disadvantages and tracking performance of maneuvering targetsand this problem leads to the need for data fusion and its multi-level structure. Then createkinematic model of indoor maneuvering target, and explore common target location trackingalgorithms, including least squares and extended Kalman filter algorithm, and proposes ahybrid optimization of both algorithms. Then elaborate the general procedure of data fusionalgorithm, with the concept of merit-based data, parameters association and data integration.This paper also focuses on research and analysis of existing data fusion algorithm, andproposes an adaptive data fusion algorithm. Finally by building a set of actual multi-source indoor positioning system, explore the structure and modules of an experimental systemplatform, and verify practical significance about data fusion in improving positioningaccuracy, reliability and real-time tracking and any other aspects. The application feedbackwas summarized in this paper, as well as inadequate results need to be improved. At last lookto the future, and describe directions worth continuing to explore.