摘要
该文针对GPS/DR(Global Positioning System/Dead-Reckoning)组合导航系统中存在参数不确定性问题,构建自组织状态空间模型,应用蒙特卡罗粒子滤波方法对新构建的系统模型进行滤波。并且对于在自组织模型中粒子滤波易使未知参数搜索陷入初始取样子集的问题,提出一种人工鱼群-粒子滤波算法。该算法不仅可以估计系统状态,而且还能使未知参数的取样分布向真实参数分布"移动",最终辨识出未知参数的真值。仿真结果表明该方法的有效性。
This paper propses a filtering method for GPS/DR (Global Positioning System/Dead-Reckoning) integrated navigation system with unknown parameters. This method firstly structures a self-organizing state space model, and then estimates the state vector by using Monte Carlo filtering method for this new system model. Because particle filter is easy to make a search of the unknown parameters into a subset of the initial sampling for the self-organization model an artificial fish swarm-partical filter algorithm is put forward. The algorithm not only can estimate the system state, but also can make the sampling distribution of the unknown parameters move to the true parameter distribution. Ultimately, the true value of the unknown parameters are identified. The simuliation results show the effectiveness of the proposed method.
出处
《电子与信息学报》
EI
CSCD
北大核心
2013年第4期921-926,共6页
Journal of Electronics & Information Technology
基金
国家自然科学基金(61273127
61074077)
高等学校博士学科点专项科研基金(20116118110008)资助课题
关键词
GPS
DR组合导航系统
自组织状态空间模型
蒙特卡罗粒子滤波
人工鱼群算法
参数辨识
GPS/DR integrated navigation system
Self-organizing state space model
Monte Carlo partical filter
Artificial Fish Swarm Algorithm (AFSA)
Parameter identification