多传感器跟踪型数据滤波融合算法

Filtering Fusion Algorithm for Multi-sensor Target Tracking

在实际系统中 ,常用的数据融合方法是基于扩展的卡尔曼滤波算法的融合算法 ,但是这种融合算法的跟踪精度并不是很高 .通过对滤波跟踪型数据融合的研究 ,提出了基于转换测量值卡尔曼滤波算法的非线性系统中的数据融合方法 .研究表明 ,在利用激光干涉仪进行目标跟踪时 ,这种基于融合算法的集中式融合算法的跟踪性能优于分布式融合算法 ,但是 ,从仿真结果可以看出 ,两种融合算法的差别并不大 ,结果基本相同 .因此 ,在非线性系统中 。

The fusion algorithm based on extended Kalman filter in common use has the disadvantage of unsatisfactory tracking precision.So this paper puts forward an fusion algorithm in nonlinear systems based on converted measurement Kalman filter. The results of simulation proved that the result of centralized converted measurement Kalman filtering is better than the result of distributed converted measurement Kalman filtering .But the difference between the two algorithms is small. So it is concluded that in nonlinear systems distributed fusion algorithm based on converted measurement Kalman filtering can basically reconstruct a centralized fusion algorithm.