分布式水下定位网中的时延差定位算法研究与实现

发布时间：2018-06-09 00:50

本文选题：Chan + Fang　；参考：《哈尔滨工程大学》2014年硕士论文

【摘要】：分布式水声定位技术根据其定位方式的不同,可以分为有源和无源定位技术两种;根据定位算法的不同,可以选择不同的数学定位模型,其中包括双曲线定位、球面定位数学模型。无论是哪种数学定位模型,目标与水下潜标之间的斜距都是通过声音传播时间与速度的乘积得到的,目标位置估计是通过求解非线性方程组得到的。本文介绍了三种常用的基于信号到达时间差(TDOA)的定位算法:Chan算法、Fang算法和Taylor算法。通过仿真分析了不同算法的定位精度,时延差定位算法在高斯环境下,潜标数目为三个时Fang氏算法比较精确,潜标数目大于等于四个时Chan算法优于Fang算法,是因为Chan算法能利用分布式定位系统中的所有时延差测量值,可以减少随机测量误差带来的影响,获得较好的定位结果;在TDOA测量误差较小时,Chan和Taylor两种算法的定位精度比较接近,但是随着时延差误差的增大Taylor展开法要优于Chan氏算法,定位精度要高一些。当Taylor展开算法初值选择较好时,Taylor级数展开算法要比Chan氏算法的定位精度高,但当Taylor算法的初值选择不理想时,定位算法的精度就会很不好。本文综合以上几种算法总结出一种基于Taylor展开法、Fang算法和Chan氏算法的协同定位法。当潜标数目为三个时将Fang算法的仿真得到的移动目标位置作为Taylor展开算法的初始值;当潜标数目大于等于四个时,将Chan仿真得到的移动目标位置作为Taylor展开法的初始值。协同算法精度明显高于其中任何一种单独的算法。通过对定位误差形成原因的推导和分析,得出了影响定位系统的关键因素,其中包括时延估计、声学定位系统本身引起的定位误差、GPS定位系统引起的定位误差、水声信道多途所引起的误差、声速测量误差引起的定位误差以及由于目标运动对定位的影响。分析不同误差下定位区域内各个点的定位精度并对误差进行补偿,采用等效声速剖面进行时延差计算;在定位区域内,中心位置定位精度最高,定位误差起伏较小,以中心向外辐射,距离中心越远的区域定位精度越低起伏越大;结合各项技术的现状和目标系统的性能,设计了定位系统的定位算法并进行了验证。
[Abstract]:The distributed underwater acoustic positioning technology can be divided into active and passive positioning technology according to its different positioning methods, and different mathematical positioning models can be selected according to the different localization algorithms, including hyperbolic positioning, The mathematical model of spherical positioning. Regardless of the mathematical positioning model, the oblique distance between the target and the underwater submersible is obtained by the product of sound propagation time and velocity, and the target position estimation is obtained by solving the nonlinear equations. In this paper, three common localization algorithms based on time difference of arrival (TDOAA) are introduced, namely: Chan algorithm Fang algorithm and Taylor algorithm. The location accuracy of different algorithms is analyzed by simulation. In Gao Si environment, Fang's algorithm is more accurate when the number of submarklets is three, and the number of submarkage is greater than or equal to four times Chan algorithm is better than Fang algorithm. The reason is that Chan algorithm can make use of all the time-delay difference measurements in distributed positioning system, which can reduce the impact of random measurement error and obtain better localization results. When the measurement error of TDOA is small, the accuracy of two algorithms, Chan and Taylor, is close, but with the increase of delay error, Taylor expansion method is better than Chan's algorithm, and the positioning accuracy is higher. When the initial value of Taylor expansion algorithm is better, the accuracy of Taylor series expansion algorithm is higher than that of Chan's algorithm, but when the initial value selection of Taylor expansion algorithm is not ideal, the accuracy of the algorithm will be very poor. In this paper, we sum up a cooperative localization method based on Taylor expansion algorithm Fang algorithm and Chan's algorithm. When the number of submarklets is three, the moving target position obtained by Fang algorithm is taken as the initial value of Taylor expansion algorithm, and the moving target position obtained by Chan simulation is regarded as the initial value of Taylor expansion method when the number of submarkmarks is greater than or equal to four. The precision of cooperative algorithm is obviously higher than that of any single algorithm. Through the derivation and analysis of the reasons for the formation of the positioning error, the key factors affecting the positioning system are obtained, including the time delay estimation, the positioning error caused by the acoustic positioning system and the positioning error caused by the GPS positioning system. The errors caused by multiple paths in underwater acoustic channels, the positioning errors caused by the measurement errors of sound velocity, and the effects of target motion on positioning. This paper analyzes the positioning accuracy of each point in the location area under different errors and compensates the error, and calculates the delay difference by using the equivalent sound velocity profile. In the location area, the positioning accuracy of the center position is the highest, and the fluctuation of the positioning error is relatively small. With the center radiating, the location accuracy of the region farther away from the center is lower and more fluctuating. Combined with the present situation of each technology and the performance of the target system, the location algorithm of the positioning system is designed and verified.
【学位授予单位】：哈尔滨工程大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TB56

【参考文献】