一种改进的自适应网格交互多模型跟踪算法

A modified adaptive grid Interacting Multiple Model tracking algorithm

自适应网格交互多模算法能够根据目标的运动模式,自适应调整系统各模型参数,从而实现对目标的跟踪。但当目标机动性较大时,在其模式变换时刻,网格中心剧烈摆动,导致跟踪精度下降。提出对该算法的改进,通过对网格中心运动步长进行适当的平滑处理,以增加网格的平稳性。仿真给出了平滑因子的经验值,并表明改进的自适应网格交互多模算法能解决传统算法在目标运动模式变换时跟踪不稳定的问题,提高精度跟踪。

The adaptive grid Interacting Multiple Model algorithm can adjust model parameter adaptively according to the target＇ s mode, and thus realizes target tracking. However, when the target maneuvering is strong at the mode changing moment, the grid center will acutely swing which degrades the tracking performance. A modified algorithm is proposed here, it can improve the grid＇ s stationarity by smoothing the moving step of the grid center properly. The empiric value of the smoothing factor is provided in the simulation part. Furthermore, the simulation results indicate that the modified adaptive grid Interacting Multiple Model algorithm can improve the instability problem in the traditional algorithm when the target＇ s mode changes and provide high precise tracking.