基于纯方位测量信息的多无人机协同目标跟踪
Collaborative Target Tracking of Multiple UAVs with Bearing-only Measurements
吴桐 1李湛 2郑晓龙 1孙一勇 3张洪梅4
作者信息
- 1. 哈尔滨工业大学 航天学院智能控制与系统研究所·哈尔滨·150001
- 2. 哈尔滨工业大学 航天学院智能控制与系统研究所·哈尔滨·150001;鹏城实验室数学与理论部·深圳·518055
- 3. 北京理工大学 宇航学院·北京·100081
- 4. 航天恒星科技有限公司·北京·100086
- 折叠
摘要
针对多无人机编队基于纯方位观测信息对目标进行协同定位和跟踪时的异步信息融合和多平台协同定位问题,提出了一种线性化的伪测量模型.通过几何推导建立起目标位置与测量位置的线性关系,在假设目标恒加速运动情况下,利用卡尔曼滤波实现了多平台异步方位信息的数据融合,并对目标运动状态进行准确估计.同时为该状态估计器设计了一种基于事件触发的状态更新策略.为实现分布式计算状态估计,基于协方差交叉原理设计了异步局部状态估计结果全局融合方法.最后通过仿真实验验证了所提出算法的有效性,并证明提出的异步数据状态估计方法与基于伪同步的扩展卡尔曼滤波方法相比在目标速度持续变化条件下具有更好的收敛速度和精度.
Abstract
Aiming at the problems of asynchronous information fusion and multi-platform cooperative positioning when multi-UAV formation based on pure azimuth observation information for cooperative positioning and tracking the target,a linear pseudo-measurement model is proposed. The linear relationship between the target position and the measurement po-sition is established through geometric derivation. Under the assumption of constant acceleration motion of the target,the data fusion of multi-platform asynchronous azimuth information is realized by using the Kalman filter,and the target motion state is accurately estimated. At the same time,a state update strategy based on event triggering is designed for the state estimator. To realize the state estimation of distributed computing,a global fusion method of asynchronous local state estimation results is designed according to the principle of covariance crossing. Finally,the effectiveness of the proposed al-gorithm is verified by simulation experiments,and it is proved that the proposed asynchronous data state estimation method has better convergence speed and accuracy than the pseudo-synchronous extended Kalman filter method under the condition of continuous changes in target velocity.
关键词
纯方位/异步数据融合/卡尔曼滤波/协同定位/状态估计/轨迹跟踪Key words
bearing-only/asynchronous data fusion/Kalman filtering/cooperative localization/state estimation/trajectory tracking引用本文复制引用
基金项目
国家自然科学基金面上项目(62273122)
黑龙江省自然科学基金优秀青年项目(YQ2023F009)
出版年
2024
国家科技期刊平台
NSTL
万方数据