| 引用本文: | 胡振涛,张勇,刘先省.基于量测迭代更新集合卡尔曼滤波的机动目标跟踪算法[J].控制理论与应用,2014,31(11):1517~1523.[点击复制] |
| HU Zhen-tao,ZHANG Yong,LIU Xian-xing.Maneuvering target tracking algorithm based on ensemble Kalman filter with observation iterated update[J].Control Theory and Technology,2014,31(11):1517~1523.[点击复制] |
|
| 基于量测迭代更新集合卡尔曼滤波的机动目标跟踪算法 |
| Maneuvering target tracking algorithm based on ensemble Kalman filter with observation iterated update |
| 摘要点击 3457 全文点击 1216 投稿时间：2014-04-16 修订日期：2014-07-11 |
| 查看全文 查看/发表评论 下载PDF阅读器 |
| DOI编号 10.7641/CTA.2014.40325 |
| 2014,31(11):1517-1523 |
| 中文关键词 机动目标跟踪 非线性滤波 集合卡尔曼滤波 交互式多模型 |
| 英文关键词 maneuvering target tracking nonlinear filter ensemble Kalman filter interacting multiple model |
| 基金项目 国家自然科学基金资助项目(61300214); 河南省高校科技创新团队支持计划资助项目(13IRTSTHN021); 河南省基础与前沿技术研究计划资助项目(132300410148); 河南省教育厅科学技术研究重点资助项目(13A413066); 中国博士后基金资助项目(2014M551999); 河南省博士后基金资助项目(2013029); 河南大学教学改革重点资助项目(HDXJJG2013--07); 河南大学优秀青年培育基金资助项目(0000A40366); 河南省社会科学规划资助项目(2011FZH005); 河南省青年骨干教师资助计划(2013GGJS$-$026). |
|
| 中文摘要 |
| 在机动目标跟踪中, 用于模型辨识和状态估计的非线性滤波器的合理选择和优化是提升滤波精度的关键. 融合量测迭代更新集合卡尔曼滤波和交互式多模型(interacting multiple models, IMM)方法, 本文提出了基于量测迭代更新集合卡尔曼滤波的机动目标跟踪算法. 通过迭代更新思想的引入构建了一种量测迭代更新下集合卡尔曼滤波的实现结构, 并将其作为IMM的模型滤波器实现对于目标运动模式和状态的辨识与估计. 针对算法结合过程中滤波精度和计算量的平衡, 设计了用于输入交互环节的状态估计样本, 同时简化输入交互环节和输出交互环节中滤波误差协方差矩阵的交互过程. 理论分析和仿真结果验证了算法的可行性和有效性. |
| 英文摘要 |
| The reasonable selection and optimization of nonlinear filter used in maneuvering target tracking is the key to realize the identification model and the estimation of state. Combining with ensemble Kalman filter with observation iterated update and interacting multiple models, a novel maneuvering target tracking algorithm based on ensemble Kalman filter with observation iterated update is proposed. Firstly, according to the mechanism of iterated observation update, the realization framework of ensemble Kalman filter with observation iterated update is constructed. And then, the improved method is taken as model filter in interacting multiple models to identify and estimate the motion mode and target state. Aiming to the balance between filtering precision and calculated amount in the combination of two algorithms, state estimation samples are designed to use in interactive input, and simultaneously the interactive processes of estimation error covariance matrix in interactive input and interactive output are simplified. The theoretical analysis and experimental results show the feasibility and validity of the new algorithm. |