H∞鲁棒滤波在机动目标跟踪中的应用

在交互多模型中通常使用的卡尔曼滤波器中，引入广义H∞鲁棒滤波器，以一定的精度为代价，换取满意的鲁棒性能。H∞鲁棒滤波算法可以分解为卡尔曼滤波和鲁棒化两个环节，从而形成一种基于增益失调因子的结构化分解算法。、为验证算法的有效性，进行了Monte Carlo仿真。仿真结果表明，本文算法跟踪复杂机动目标时跟踪性能有较大提高，有很好的可实现性.

Application of the H∞ Robust Filter in Maneuvering Target Tracking

A generalized H∞ robust filtering algorithm was introduced instead of Kalman filtering using in IMM algorithm. H∞ filter achieves satisfying robustness at the cost of some estimation precision. A suit of ＂structuring decomposed algorithm＂ was established when H∞ robust filtering algorithm was decomposed in two parts： Kalman filtering and making robustness. In order to verify the validity of this algorithm, Monte Carlo simulation was performed. The simulation results indicate that the tracking performance was proved when tracking complex maneuvering target, and the algorithm is very practicable.