厚尾噪声条件下星凸形扩展目标student’s t滤波器

Star-convex extended target student’s t filter for heavy tailed noise

针对非线性非高斯厚尾噪声条件下的不规则形状的扩展目标跟踪问题,提出了基于随机超曲面模型(RHM)的星凸形扩展目标student’s t滤波算法.首先,在带有非高斯厚尾过程噪声和厚尾量测噪声的系统中,基于student’s t分布给出鲁棒student’s t滤波器.其次,利用随机超曲面模型描述任意星凸形扩展目标的量测源分布,提出带厚尾噪声的星凸形扩展目标student’s t滤波器.最后通过仿真验证了所提方法的正确性和有效性.

Aiming at the problem of irregular shape extended target tracking with nonlinear non-Gaussian heavy tailed noise,this paper proposes a star-convex extended target nonlinear student’s t filter algorithm based on a random hypersurface model.First,in a system with non-Gaussian heavy tailed process and measurement noise,a robust student’s t based nonlinear filter is given based on the student’s t distribution.Then,random hypersurface model(RHM)is used to describe the measurement source distribution of any star-convex extended target,and star-convex extended target student’s t filter is proposed.Finally,numerical simulation example shows the validity and feasibility of the proposed method.