基于Matlab仿真的当前统计模型的参数校正

The Parameters Adjustment of Current Statistical Model Based on Matlab Simulation

当前统计模型能真实地反映目标机动范围和强度的变化,是目前较好的实用模型。大量实验表明该算法在跟踪机动目标时具有良好的跟踪结果。然而实验中也发现该算法在跟踪具有加速度的目标机动情况时,其速度与加速度估计的动态时延明显位置误差较大,因此不能很好地实时反映目标的机动情况。因此需要进行新的调整参数的设定与比较,使其克服以上的缺点,文章借鉴强跟踪滤波器,在滤波器状态预测协方差矩阵中引入了加权因子并利用M atlab仿真技术,针对当前统计模型中对动态时延影响比较大的几个重要参数,进行了仿真对比和调整。跟踪结果表明:动态时延明显减小,位置误差大幅下降,达到了比较理想的跟踪效果。

The current statistical model is a good practical model which can reflect the scope and intensity change of the maneuvering target. Large numbers of experiments show that the algorithm does well in tracking maneuvering targets , However, the experiment also found that the dynamic delay of velocity and acceleration is significantly great and the position error is large in tracking a maneuvering target. Hence adjustments of the parameters are needed to overcome these shortcomings. Referred to strongly tracking filter, weighting factors are introduced for the filter state prediction covariance matrix. And several important parameters which impact most are adjusted according to simulation results. Simulation results show that： Dynamic delay and the position error is significantly reduced, which reaches more satisfying tracking performance.