采用序贯滤波的红外/雷达机动目标跟踪

Infrared/radar maneuvering target tracking via sequence filtering

当机动目标状态是非平稳和非线性时,红外传感器和雷达的目标状态方程和量测方程都是非线性和非高斯的,为了解决经典跟踪算法的残差较大或发散的问题,提出一种新的融合跟踪算法:在对红外传感器和雷达的量测数据进行时间对准和同步融合后,将融合后的量测数据送入重抽样粒子滤波器进行处理以预测和跟踪机动目标.最后给出了一个仿真跟踪实例,并与同类多雷达跟踪的效果进行了比较,说明了异类融合跟踪优于同类多雷达融合跟踪.

When the trace of the infrared small weak target is nonlinear and non-stationary which are always appear in practice, the state equation and measurement equation are nonlinear-non-Gaussian and it is hard to get the solution using traditional algorithm. A new maneuvering target tracking algorithm fusing the measurement of infrared sensor and radar sensor is proposed in this paper. After the process of infrared/radar time registration and space registration, the measurement data after fusion is send into the resampling particle filter to track the target. A simulation example is also given and compared with a two-radar target tracking, showing the advantage of the fusion tracking over the two-radar target tracking.