边缘化迭代容积卡尔曼滤波的单站无源定位算法

针对单站无源定位可观测性弱、观测噪声大而导致的定位精度低、收敛速度慢等问题，提出了一种边缘化迭代容积卡尔曼滤波算法。该算法采用基于似然增加的迭代策略，不需要设置判决门限，且保证了算法的全局收敛性。同时，其充分考虑状态向量与观测噪声之间的互协方差，将状态向量扩维，构造条件线性模型并进行边缘化滤波，不仅提高了算法的定位精度以及收敛速度，还减少了扩维后所需的采样点，提高了算法的运算效率。仿真结果表明，新算法改善了单站无源定位的定位精度以及收敛速度。 

Single Observer Passive Location Algorithm Based on Marginalized Iterated Cubature Kalman Filter

Because of the low observability and the high noise in single observer passive location, the performance of the positioning accuracy and convergence velocity was poor. A novel marginalized iterated cubature kalman filter was presented. The iteration strategy based on the likelihood increase was adopted. The global convergence of the algorithm was ensured, without the difficult choice of the judgment threshold. Meanwhile, the cross-covariance between the state and the measurement noise was taken into account. Then the state vector was augmented by the measurement noise. Based on the conditionally linear model, the marginalized filtering was proceeded. The positioning accuracy and the convergence velocity was improved. And the computational burden was reduced, because the less sigma points were needed in spite of the augmented state vector. Simulation results indicate that the novel algorithm improved the performance of the positioning accuracy and convergence velocity in single observer passive location.