• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>IEEE Transactions on Aerospace and Electronic Systems >Efficient particle filtering for road-constrained target tracking
【24h】

Efficient particle filtering for road-constrained target tracking

机译：高效的粒子滤波技术，用于道路受限的目标跟踪

获取原文
获取原文并翻译 | 示例
           
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

The variable-structure multiple-model particle filtering approach for state estimation of road-constrained targets is addressed. The multiple models are designed to account for target maneuvers including "move-stop-move" and motion ambiguity at an intersection; the time-varying active model sets are adaptively selected based on target state and local terrain condition. The hybrid state space is partitioned into the mode subspace and the target subspace. The mode state is estimated based on random sampling; the target state as well as the relevant likelihood function associated with a mode sample sequence is approximated as Gaussian distribution, of which the conditional mean and covariance are deterministically computed using a nonlinear Kalman filter which accounts for road constraints in its update. The importance function for the sampling of the mode state approximates the optimal importance function under the same Gaussian assumption of the target state.
机译：提出了一种用于道路受限目标状态估计的可变结构多模型粒子滤波方法。多个模型旨在解决目标机动问题，包括交叉路口的“移动-停止-移动”和运动歧义；基于目标状态和当地地形条件自适应选择时变活动模型集。混合状态空间被划分为模式子空间和目标子空间。模式状态是基于随机采样估计的；目标状态以及与模式样本序列相关的相关似然函数近似为高斯分布，其条件均值和协方差是使用非线性卡尔曼滤波器确定性地计算的，该非线性卡尔曼滤波器考虑了其更新中的道路约束。模式状态采样的重要性函数在目标状态的相同高斯假设下近似最佳最优重要性函数。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号