Research on real-time elimination of ultra-wideband radar ranging abnormal value data

Abstract. For indoor positioning, ultra-wideband (UWB) radar comes to the forefront due its strong penetration, anti-jamming, and high-precision ranging abilities. However, complex environment disorder of obstacles, problems diffraction, instability caused by UWB signals also emerge, which make it difficult predict noise leads a great impact on accuracy stability measurement data in short term. Therefore, abnormal value migration positioning trajectory occurred real-time positioning. To eliminate this phenomenon provide more accurate results, values need be removed. It is not accurately based large number data, but still problem ensure system using small time data. Thus, paper focuses experimental analysis UWB-based system. By repeatedly measuring range , can obtained. Using massive train linear regression models, we get parameter model measured with radar. Based Gaussian function outlier detection, are eliminated, putting new into trained us, error reduced nearly 50 % compared peak mean errors general.