基于新息偏差的观测奇异值自适应滤波算法

The Self-Adaptive Filtering Algorithm of Outlier Observations Based on Innovation Deviation

针对在机动目标跟踪过程中出现的观测奇异值,提出了一种对观测奇异值进行自适应滤波的算法。依据多传感器数据形成的对目标状态参数的正确描述和测量数据集合主体的变化趋势,利用新息偏差(innovation deviation)实时、准确地识别测量数据中的奇异点,并进行自适应滤波。仿真结果表明,在分布式多传感器目标跟踪的航迹融合处理中,该算法可以快速、有效地解决奇异值自适应滤波的问题。

Aimming at the outliers coming from the process of tracking the maneuver target,a self-adaptive filtering algorithm is presented.Therefore,the outliers correction is of great importance in the improving of the accuracy of target-tracking.In distributed multi-sensors target-tracking system,the outliers are recognized and eliminated at real-time by using innovation deviation according to the trend of the principal target tracking data acquisition by multi-sensors.Simulation results show that the algorithm can effectively filters outliers quickly in the distributed multi-sensors target tracking system.