提出了基于Tikhonov正则化理论的多站时差定位新方法.将基于拟最优原则的Tikhonov正则化方法应用于经过线性化处理后的定位方程组的求解,这种方法的优势之一在于它不需要任何关于噪声分布的信息.给出了有关正则解收敛性的理论结果,并介绍了正则化定位的计算方法.在计算机仿真部分,以普通最小二乘定位算法为参照,给出了在不同误差条件下两种算法的定位结果.数据结果显示,正则化定位算法的定位精度相对较高,且抗噪声干扰的能力也较强,这证明了正则化定位算法的可行性。
A novel multi-station TDOA location method based on Tikhonov regularization theory was proposed. The regularization method, with the advantage that it needed no assumptions on the distribution of noises, was adopted to solve the linearized location equations with a quasi-optimal principle. The convergence of the regularization solution was theoretically confirmed, and then the regularization location procedure was introduced. In simulation part, regularization location method was compared with the ordinary least square (OLS) location method under different noise conditions. The regularization location shows a higher estimate precision and better ability of resisting noises than OLS method. Conclusions can be drawn that regularization location method is practicable.