首页|基于CKF-SLAM改进的无人水下航行器动态目标跟踪算法研究

基于CKF-SLAM改进的无人水下航行器动态目标跟踪算法研究

扫码查看
原文链接
  • 万方数据
  • 维普
针对容积卡尔曼滤波(cubature Kalman filter,CKF)同步定位与建图(simultaneous localiza-tion and mapping,SLAM)算法在动态目标跟踪(object tracking,OT)的应用中,存在算法实时性不高、计算复杂以及对动态目标物跟踪精度较低的问题,提出基于平方根容积卡尔曼滤波SLAM的无人水下航行器(unmanned underwater Vehicle,UUV)目标跟踪算法(SRCKF-SLAM-OT).该算法将CKF-SLAM-OT中复杂的计算部分,利用3 阶容积准则选取一组相同权值的容积点来近似计算,再利用数值积分法计算非线性方程模型的后验状态估计平均值和方差,并对协方差矩阵的平方根因子进行更新.仿真结果表明:SRCKF-SLAM-OT算法简化了计算量和改善了数值精度,提高了UUV在未知水下环境中自身定位的精度和动态目标物跟踪的能力.
Research on improved dynamic object tracking algorithm for unmanned underwater vehicle based on CKF-SLAM
For the application of the Cubature Kalman Filter(CKF)Simultaneous Localization and Mapping(SLAM)algorithm in the dynamic Object Tracking(OT),there are the problems of low real-time performance,complex calculation and for the problem of low tracking accuracy of dynamic objects,An object tracking algorithm for Unmanned Underwater Vehicle(UUV)based on square root cubature Kalman filter SLAM(SRCKF-SLAM-OT)is proposed.The algorithm uses the third-order cubature criterion to select a set of volume points with the same weight for approximate calculation of the complex calculation part in CKF-SLAM-OT,and then uses the nu-merical integration method to calculate the posterior state estimated average value and sum of the nonlinear equation model.variance,and update the square root factor of the covariance matrix.The simulation results show that the SRCKF-SLAM-OT algorithm simplifies the calculation and improves the numerical accuracy,and improves the UUV's self-positioning accuracy and dynamic object tracking ability in unknown underwater environments.

dynamic object trackingcubature kalman filtersimultaneous localization and mappingsquare-root cubature kalman filterunmanned underwater vehicle

都立立、邢传玺、万志良、李聪颖

展开 >

云南民族大学 电气信息工程学院,云南 昆明 650500

动态目标跟踪 容积卡尔曼滤波 同步定位与建图 平方根容积卡尔曼滤波 无人水下航行器

国家自然基金云南省基础研究专项面上项目云南省高校信息与通信安全灾备重点实验室

61761048202101AT070132650500

2024

10.3969/j.issn.1672-8513.2024.01.013

云南民族大学学报(自然科学版)
云南民族大学

云南民族大学学报(自然科学版)

CSTPCD
影响因子：0.381
ISSN：1672-8513
年,卷(期)：2024.33(1)
  • 1
  • 14