基于迭代CDKF的单站无源定位算法

Iterated CDKF for single observer passive location

针对单站无源定位系统存在滤波稳定性差、收敛速度慢和定位精度差等问题,提出一种迭代中心差分卡尔曼滤波算法.在迭代判决准则的约束下,重复利用观测信息对状态向量和误差协方差矩阵进行迭代估计使其更趋向真实值,同时用Levenberg-Marquardt优化方法调整预测误差协方差阵,保证算法的全局收敛.仿真结果表明,在不同参数测量精度条件下,该算法稳定性、收敛速度和定位精度较好.

A new novel iterated central difference Kalman filter algorithm was presented to solve the issues of the filter in the single observer passive location system,such as bad stability,slow convergence speed and poor locating accuracy.In the iterative sentencing guideline constraints,the new algorithm can generate more accurate and reasonable estimation of the state vector and error covariance matrix by reusing the observation information.The Levenberg-Marquardt optimization method is used to adjust the forecast error covariance matrix.This ensures the global convergence of the algorithm.Simulation results show that the new algorithm has stabler performance,larger convergence speed and higher locating accuracy in different scenarios.