摘要
提出了一种在LOS(视距)环境下基于神经网络的UWB精确定位算法,通过BP神经网络算法对数据进行大量训练,得到最优化的神经元权系数,形成神经网络,从而实现TDOA(到达时间差)定位的解算。仿真结果表明,在UWB的通信环境下,与经典的Taylor算法和Chan算法比,方法在实时处理要求下定位性能是最优的,在计算复杂度代价与定位精度方面取得折中,可满足实际定位应用中低功耗、低复杂度和快速高精度定位的要求。
A UWB precise positioning algorithm based on neural network in LOS(Line of Sight) environment is proposed.In order to get the optimized nerve cell coefficients,plenty of training is done by BP neural network,so that the resolution of TDOA positioning is achieved.Simulation results show that this method is the optimized method compared to Taylor and Chan algorithm in UWB communication environments,since it can get a balance between the calculation costs and positioning precision,what′s more,it can satisfy the demands of low power consumption,low complexity and fast high precise positioning.
作者
孙晔
肖竹
李小蓓
向新
SUN Ye;XIAO Zhu;LI Xiao-bei;XIANG Xin(Aeronautics Engineering College,Air Force Engineering University,Xi'an 710038,China;College of Computer and Communication,Hunan University,Changsha 410082,China;Institute of Information and Navigation,Air Force Engineering University,Xi'an 710077,China)
出处
《航空计算技术》
2019年第2期6-10,共5页
Aeronautical Computing Technique
基金
国家自然科学基金项目资助(61701524)
关键词
超宽带
TDOA
定位
神经网络
ultra wideband
TDOA
positioning
neural network
