基于QMC采样的GMPHD分布式融合方法

发布时间：2018-03-17 14:45

本文选题：高斯混合模型　切入点：分布式融合　出处：《系统工程与电子技术》2017年08期 　论文类型：期刊论文

【摘要】：针对高斯混合概率假设密度分布式融合过程中高斯分量数随时间急剧增长的问题,给出了一种适用于融合过程不同阶段的两级分量混合约简算法,最大程度地减少了信息的损失。针对高斯混合概率假设密度协方差交叉融合算法中高斯混合模型求幂运算后不再服从高斯混合分布的问题,提出了一种基于拟蒙特卡罗采样的等价求解方法。仿真实验表明,所提的改进算法在保证融合计算有效性和可行性的同时提高了融合精度。
[Abstract]:In order to solve the problem of the rapid increase of Gao Si component number with time in Gao Si's mixed probability assumption density distributed fusion process, a two-stage hybrid reduction algorithm for different stages of fusion process is presented in this paper. The loss of information is reduced to the greatest extent. For the question of the mixed distribution of Gao Si mixed model after power operation in Gao Si's mixed probability assumption density covariance cross fusion algorithm, An equivalent solution method based on quasi Monte Carlo sampling is proposed. The simulation results show that the proposed improved algorithm not only ensures the validity and feasibility of fusion computation, but also improves the fusion accuracy.
【作者单位】：西安测绘总站;空军工程大学信息与导航学院;
【基金】：国家自然科学基金(61571458) CEMEE国家重点实验室开放基金项目(2014K0304B)资助课题
【分类号】：TN713

【相似文献】