一种基于自适应去偏转换滤波的机动目标跟踪

针对于机动目标的跟踪问题,提出了一种基于交互式多模型的自适应去偏转换卡尔曼滤波器.该算法利用交互多模型算法来完成不同跟踪模型的相互切换;根据自适应去偏转换测量卡尔曼滤波算法来推导跟踪目标状态,同时自适应因子可以确保不正常测量时的鲁棒性.与传统的去偏转换卡尔曼滤波算法对比,该算法可以很好地改善所获量测信息在雷达被干扰时的目标跟踪精度.仿真结果表明了算法的有效性和可行性,且跟踪精度相对传统的去偏转换卡尔曼滤波算法减少9.38％的位置误差.     

Numerical errors of the volume‐of‐fluid interface tracking algorithm

One of the important limitations of the interface tracking algorithms is that they can be used only as long as the local computational grid density allows surface tracking. In a dispersed flow, where the dimensions of the particular fluid parts are comparable or smaller than the grid spacing, several numerical and reconstruction errors become considerable. In this paper the analysis of the interface tracking errors is performed for the volume‐of‐fluid method with the least squares volume of fluid interface reconstruction algorithm. A few simple two‐fluid benchmarks are proposed for the investigation of the interface tracking grid dependence. The expression based on the gradient of the volume fraction variable is introduced for the estimation of the reconstruction correctness and can be used for the activation of an adaptive mesh refinement algorithm. Copyright © 2002 John Wiley & Sons, Ltd.