论文部分内容阅读
提出了改进型约束总体最小二乘多目标定位算法.首先引入辅助变量将非线性定位方程转化为伪线性方程;然后利用两步最小二乘法估计目标的初始位置,依据目标初始位置重新选择参考传感器;最后考虑伪线性方程中所有系数矩阵的噪声,采用拉格朗日乘子技术求解约束条件,利用拟牛顿算法迭代公式得到精确解.仿真结果证明了理论分析的正确性和可行性,所提算法能够达到克拉美罗下界,具有较强的鲁棒性和精确的定位性能.
An improved constrained global least squares multi-target localization algorithm is proposed.Firstly, auxiliary variables are introduced to transform the nonlinear positioning equation into pseudo-linear equations, then the two-step least square method is used to estimate the initial position of the target, and the reference sensor is re-selected according to the target initial position Finally, the noise of all the coefficient matrices in the pseudo-linear equation is considered. The Lagrange multiplier technique is used to solve the constraint and the exact solution is obtained by using the iterative formula of quasi-Newton algorithm. The simulation results prove the correctness and feasibility of the theoretical analysis. The algorithm can reach the lower limit of clarithromycin and has strong robustness and precise positioning performance.