基于H_∞滤波的不同目标机动模型的跟踪特性研究

An Investigation on Different Maneuvering Models in High-g Target Tracking Using H∞Filtering

文中研究了五种常见的目标机动模型下H∞滤波用于机动目标跟踪的性能。在目标位置信息可测条件下针对机动目标进行了H∞跟踪滤波仿真计算,以速度估计均方差为指标,分析了不同目标机动模型的H∞跟踪滤波精度。结果表明,基于维纳模型的H∞滤波进行机动目标跟踪精度最高,采用常加速度模型或二阶马尔可夫模型也是一种合适的选择。

H∞filtering technique has been extensively applied in maneuvering-target tacking problems. Although H∞filter is robust to the uncertainties in target dynamic models and other disturbances existing in a system, the target model selection is still very important when high tracking performance is required for high-g maneuvering targets, especially in precision guidance systems. This paper focus mainly on the performance of H∞filtering under different target models. Five target models are considered and the tracking precisions are assessed and compared with only target position measurements. The simulation results show that when H∞filter is used, the best selection of target models is the wiener model, and constant acceleration model or 2-order Markov models is secondary choice.