基于多假设的航迹关联方法研究  

Research on track association algorithm based on multi-hypothesis thought

作　　者：文云峰[1] 石章松[1] 王芳[2] 

机构地区：[1]海军工程大学电子工程学院 [2]中国人民解放军91746部队

出　　处：《舰船科学技术》2011年第2期116-120,共5页Ship Science and Technology

摘　　要：提出了一种基于多假设思想(MHT)的航迹关联方法,将多平台多目标航迹关联问题转化为假设树的构建和管理问题,利用树形结构特征进行快速剪枝,极大地减少了错误假设的冗余计算量。先通过航迹粗精两级关联,形成关联假设,然后设计基于关联假设的航迹关联假设树,根据关联假设历史累积的关联质量搜索最优关联解。这种关联能充分利用航迹的各种特征信息和历史累积信息来提高关联的正确性和实时性。仿真计算表明,该算法具有较高的正确关联率,在复杂环境下达到了良好的效果。A track association algorithm based on multi-hypothesis thought is presented, which translate the multi-platform and multi-target track association algorithm into the construct and manage of hypothesis trees, and using multiple trees method to reduce the redundant computing by pruning the wrong hypothesis tracks. At the first ,we can form the association hypothesis by the coarse association and the fine association, then design the association hypothesis tree based on the corresponding relationship of the association hypothesis, at last, the optimal solution is searched by history cumulate association quality of association hypothesis. This association can improve the correctness and real-time performance by using character and cumulate information of the track. As being validated by simulation results, the correct association is of high probability ,and get good effects in complex circumstance

关 键 词：航迹关联 多假设 多传感器 融合 

分 类 号：TP911.72[自动化与计算机技术]