journal6 ›› 2005, Vol. 26 ›› Issue (4): 15-20.

• 物理与电子 • 上一篇    下一篇

分布式不敏卡尔曼滤波状态估计技术

  

  1. (海军航空工程学院信息融合技术研究所,山东 烟台 264001)
  • 出版日期:2005-10-15 发布日期:2012-09-17
  • 作者简介:熊伟(1977-),男,江西南昌人,海军航空工程学院信息融合技术研究所讲师,工学博士,主要从事多传感器信息融合和多目标跟踪技术研究.
  • 基金资助:

    国家自然科学基金资助项目(60172033);全国优秀博士论文作者专项基金项目(2000036)和高校骨干教师基金资助项目(3240)

Distributed  State Estimation Based on Unscented Kalman Filter

  1. (Research Institute of Information Fusion,Naval Aeronautical Engineering Institute,Yantai 264001,Shandong China)
  • Online:2005-10-15 Published:2012-09-17

摘要:在许多实际的分布式多传感器系统中,系统的动态或传感器的观测方程是非线性的.解决分布式多传感器非线性系统的状态估计问题,通常采用的一种方法是分布式扩展卡尔曼滤波.但由于模型的线性化误差,EKF的滤波效果在很多情况下并不能令人满意.另外,在许多实际应用中,模型的线性化过程比较繁杂,而且也不容易得到.为了有效解决分布式多传感器非线性系统的状态估计问题,提出了一种基于不敏卡尔曼滤波的状态估计技术.不敏卡尔曼滤波是最近提出的一种新的非线性滤波方法.由于不需要对非线性系统进行线性化,不敏卡尔曼滤波可以很容易地应用于非线性系统的状态估计,并且其性能也要优于扩展卡尔曼滤波.仿真结果说明分布式不敏卡尔曼滤波方法的性能要优于分布式扩展卡尔曼滤波方法.

关键词: 分布式, 非线性, 多传感器, 状态估计, 不敏卡尔曼滤波

Abstract: In most applications of distributed multisensor system,the system dynamics or observation equations are nonlinear.The Extended Kalman Filter (EKF) is the most common approach for distributed multisensor nonlinear state estimation.It is well known that the performance of the EKF may not be always good due to the linearization error.In addition,the derivation of the Jacobian matrices is nontrivial in most applications and often lead to implementation difficulties.To address distributed multisensor nonlinear state estimation problems,the paper proposes a novel state estimation method based on the unscented kalman filter.The Unscented Kalman Filter (UKF) is a novel method for nonlinear filtering.The UKF can be more easily applied for nonlinear state estimation,because the UKF does not approximate the nonlinear state and measurement equations.Moreover,The UKF can get more accurate estimation than the EKF.At last,a Monte Carlo simulation is used to analyze the performance of the method.The results of the simulation prove that the new method will get more accurate estimation than the distributed Extended Kalman Filter.

Key words: distributed, nonlinear, multi-sensor, state estimation, unscented kalman filter

公众号 电子书橱 超星期刊 手机浏览 在线QQ