A hybrid GPS/TDOA(time difference of arrival algorithm) is proposed to realize location when neither GPS nor TDOA location method has positioning ability. The algorithm first makes differential operation between GPS pseudo-range equations to eliminate the constant terminal clock bias and gets GPS hyperbola equations. Then the GPS hyperbola equations are combined with TDOA hyperbola equations from cellular networks. The equations are then linearized by Taylor Series method, and the terminal location is iteratively estimated with least squares algorithm based on the linear equations. Field experiments show that more GPS pseudo-ranges will improve the location accuracy, but more TDOA measurements will degrade the location accuracy because of the big NLOS error in TDOA. It is concluded that GPS pseudo-ranges need be fully used when implementing the proposed algorithm, and TDOA measurements should be less used after ensuring location determined. In the experiments, the maximal error is 231.52 m and the root mean squared error under the best condition is 33.80 m. Such accuracy can satisfy the location applications when the precision requirement is not high but position information is indeed needed.%提出了一种混合GPS/TDOA的泰勒级数-最小二乘算法,在GPS和TDOA定位方法都无法独立定位时实现了定位.算法先将GPS伪距观测方程做单差运算、消除钟差,然后与TDOA双曲线方程组合,一起进行泰勒级数展开得到组合后的线性方程组,再利用最小二乘的方法迭代估计出终端位置.验证该算法的实验结果表明,GPS伪距观测量越多时,定位精度越高;相反,由于TDOA观测量NLOS误差较大,TDOA观测量越多时,精度越差.因此实施该算法时,GPS伪距观测量作为输入应全部采用;在保证能够定位的前提下,应尽量少用TDOA观测量.试验所得混合定位结果的最大误差为231.52 m,最佳条件下所得均方根误差为33.80m,能够满足某些精度要求不高,而又必需位置信息的定位应用需求.
展开▼
