移动终端定位的关键算法研究

发布时间：2018-05-08 23:31

本文选题：TDOA + 几何精度因子　；参考：《华南理工大学》2015年博士论文

【摘要】：随着电子、互联网和移动通信技术的不断发展,以位置信息为核心的智能交通信息工程有了在深度和广度上不断延伸的技术支撑,基于位置的服务(Location Based Service,LBS)在公众交通出行中的需求也越来越广泛和明确。同时各国政府为了提高政府管理效能,对基于位置的安全管理提出了强制性要求,这进一步带动了位置服务在更多行业和国民生活领域的广泛应用。因此,基于位置的服务体现出巨大的应用前景和商业价值,准确估计移动终端位置的重要性和必要性日益显露,引起了国内外的广泛关注和研究。本文对移动终端定位的研究现状和存在问题进行了分析,围绕提高移动终端位置估算精度为核心进行研究,提出如下几个创新点:(1)分析了几种经典的基于到达时间差(Time Difference of Arrival,TDOA)坐标解算方法,并在无线局域网(Wireless Local Area Networks,WLAN)环境下进行了仿真,从适用性、收敛性和精度几个方面对各算法做了对比分析。结合最速下降法适应性强、在初始值误差较大时收敛速度快的特点,以及Levenberg-Marquardt迭代法在估计值接近真实值时收敛速度快的特点,本文提出一个融合二者优点的混合定位法,通过WLAN环境下手持终端的数据采样和仿真分析,验证了本算法具有较高的定位精度、良好的适用性和健壮性。(2)对无线移动定位系统三种典型布局进行了详细的几何精度因子(Geometric Dilution of Precision,GDOP)推算和仿真分析,分别是车辆终端位于圆形均匀排列的中央、圆周任意位置、对称线任意位置,推测出GDOP值随终端、基站的位置关系呈现出一定的变化规律,该理论对工程应用中基站的布局实施具有良好的指导意义。提出一种在基站信号辐射距离和基站位置关系的条件约束下,采用障碍函数法和模拟退火法进行最佳布局解算的优化模型,通过车辆终端数据采样和系统仿真,本布局优化模型能够有效降低定位区域的GDOP值,达到总体上提高区域内车辆终端定位精度的效果。(3)研究了几种典型的信号统计分布和统计特征的非视距鉴别算法和误差抑制算法。利用蜂窝环境下TDOA测量方差在非视距条件下比视距条件下大这一先验信息,提出一种建立了实际测量值方差和理想视距条件下测量值方差的对比模型和判别门限的TDOA测量值非视距误差鉴别法,并提出一种对已鉴别非视距误差进行补偿的基于TDOA泰勒级数展开法的优化算法。在蜂窝通信网络下对这两种算法进行了仿真分析,误差补偿后的泰勒级数展开法的精度相比优化之前有明显提升,证明了该非视距鉴别算法的有效性。本文还提出一种基于自适应联邦卡尔曼滤波数据融合的GPS/TDOA混合定位算法,并且在GPS伪距定位算法和蜂窝TDOA定位算法中分别采用扩展卡尔曼滤波进行非视距误差抑制,通过手持终端数据采样和系统仿真,证明了经过扩展卡尔曼滤波非视距误差抑制后,GPS和TDOA的定位误差明显下降,验证了基于GPS信号时延和蜂窝TDOA两种测量值进行融合的混合定位算法的定位精度比二者单独定位的精度要高。(4)根据GPS载波相位观测模型和整数最小二乘法的模糊度解算原理,分析了几种常用的整周模糊度解算法。针对双差模式下序贯条件最小二乘法浮点解的强相关性导致模糊度解算失败的问题,本文提出采用高斯变换法对模糊度浮点解进行降相关处理,以提高模糊度解算的准确性和整数解的搜索效率;针对基线伪距可测量以及观测历元少的情况,提出一种基于基线伪距约束的方法进一步优化序贯条件最小二乘法的算法,该算法通过改善模糊度的协因数矩阵,降低浮点解的相关性,缩小模糊度的搜索空间,最终提高模糊度解算的精度以及解算稳定性和成功率,实验仿真证明了这两种优化算法的有效性。
[Abstract]:With the continuous development of electronic, Internet and mobile communication technology, intelligent traffic information engineering with location information has a continuous extension of technical support in depth and breadth. The demand for Location Based Service (LBS) in public transportation is becoming more and more extensive and clear. In order to improve the effectiveness of government management and put forward mandatory requirements for location based security management, this further promotes the wide application of location services in more industries and in the field of national life. Therefore, location based services show great application prospects and commercial value, and the importance and necessity of accurate estimation of mobile terminal position are increasingly becoming more and more important. In this paper, the research status and existing problems of mobile terminal positioning are analyzed, and the research on improving the accuracy of mobile terminal location estimation is studied. The following innovations are put forward: (1) several classic Time Difference of Arrival (TDOA) sit is analyzed. In the environment of Wireless Local Area Networks (WLAN), the simulation is carried out. The algorithm is compared and analyzed from the aspects of applicability, convergence and precision. The fast convergence rate is fast when the initial value error is large and the Levenberg-Marquardt iteration method is estimated. In this paper, the convergence speed is fast when the value is close to the real value. In this paper, a mixed location method which combines the advantages of the two is proposed. Through the data sampling and simulation analysis of the handheld terminal in the WLAN environment, it is proved that this algorithm has high positioning accuracy, good applicability and robustness. (2) three typical layout of wireless mobile positioning system are carried out. The detailed geometric precision factor (Geometric Dilution of Precision, GDOP) calculation and simulation analysis, respectively, the vehicle terminal is located in the center of a circular and uniform arrangement, at any position of the circular circle, and at any position of the symmetrical line. It is deduced that the GDOP value varies with the terminal, the position relation of the base station has a certain change law. The layout of the base station has good guiding significance. Under the condition of the relation between the base station signal radiation distance and the base station position, the optimization model of the optimal layout solution is carried out by the barrier function method and the simulated annealing method. Through the vehicle terminal data sampling and the system simulation, this layout optimization model can effectively reduce the location area. The GDOP value can improve the accuracy of vehicle terminal positioning in the region as a whole. (3) the non sight distance identification and error suppression algorithms of several typical statistical distribution and statistical characteristics are studied. The TDOA measurement variance under the non sight distance under the cellular environment is used to establish a prior information under the sight distance. The contrast model of the actual measured value variance and the measured value variance under the ideal sight distance condition and the discriminant method of the TDOA measurement value of the discriminant threshold, and propose an optimization algorithm based on the TDOA Taylor series expansion method to compensate the identified non sight distance error. The two algorithms are simulated under the cellular communication network. Analysis, the accuracy of the Taylor series expansion method after error compensation has been improved obviously before the optimization, which proves the effectiveness of the non sight distance identification algorithm. In addition, a GPS/TDOA hybrid location algorithm based on adaptive federated Calman filter data fusion is proposed, and the GPS pseudo range location algorithm and the cellular TDOA location algorithm are adopted respectively. The extended Calman filter is used to suppress the non sight distance error. Through the data sampling and system simulation of the handheld terminal, it is proved that the positioning error of GPS and TDOA is obviously reduced after the extended Calman filter is suppressed by the non sight distance error. The location of the mixed location algorithm based on the two kinds of measurement values based on the GPS signal delay and the cellular TDOA is verified. The precision is higher than that of the two one. (4) according to the GPS carrier phase observation model and the integer least square method of ambiguity resolution, several commonly used integer ambiguity resolution algorithms are analyzed. The problem that the strong correlation of the floating-point solution of the sequential conditions of the sequential conditions in the double difference mode leads to the failure of the ambiguity resolution. In order to improve the accuracy of the ambiguity resolution and the search efficiency of the integer solution, the Gauss transform is used to improve the accuracy of the ambiguity resolution and the search efficiency of the integer solution. A method for further optimization of the sequential conditional least squares is proposed for the baseline pseudo distance measurement and the few observation epochs. The co factor matrix of fuzzy degree is improved, the correlation of floating-point solution is reduced, the search space of fuzzy degree is reduced, the accuracy of ambiguity resolution, stability and success rate are improved, and the effectiveness of the two optimization algorithms is proved by experimental simulation.

【学位授予单位】：华南理工大学
【学位级别】：博士
【学位授予年份】：2015
【分类号】：U495

【参考文献】