一种基于状态相关因子的抗差自适应滤波算法  

A Robust Adaptive Filter Method Based on State Dependent Coefficient

作　　者：高怡[1] 高社生[1] 

机构地区：[1]西北工业大学自动化学院,陕西西安710072

出　　处：《测控技术》2014年第2期111-115,共5页Measurement & Control Technology

基　　金：国家自然科学基金资助项目(61174193);陕西省自然科学基金资助项目(N9YU0001)

摘　　要：提出一种新的基于状态相关因子的抗差自适应滤波算法。采用状态相关因子将非线性系统转换为状态相关系统,建立抗差自适应滤波模型,利用等价权矩阵和自适应因子进行信息分配,从而控制动力学模型异常和观测异常对导航解的影响。特定条件下的仿真计算表明,提出的基于状态相关因子的抗差自适应滤波算法,不仅能够抑制动力学模型噪声和观测噪声干扰,而且滤波计算简单,导航精度相对优于扩展卡尔曼滤波和粒子滤波算法。A new robust adaptive filter method is presented based on state dependent coefficient. The state de- pendent coefficient is adopted to convert nonlinear system into state related system, and a robust adaptive filter model is established. Equivalent weight matrix and adaptive factor are used to assign information reasonably, which controls the influence of dynamic model abnormal and observation abnormal on navigation result. Simula- tion results of certain conditions demonstrate that the proposed filter method can not only resist interference of dynamic model noise and observation noise, but also achieve higher accuracy than extended Kalman filter and particle filter.

关 键 词：状态相关因子 抗差估计 自适应滤波 自适应因子 组合导航系统 

分 类 号：U666.1[交通运输工程—船舶及航道工程]