基于扩展H_∞滤波自适应误差配准算法  

An Adaptive Multi-Sensor Registration Algorithm Based on Extended H_∞ Filtering

作　　者：任清安[1] 钮俊清[1] 刘军伟[1] 

机构地区：[1]中国电子科技集团公司第三十八研究所,安徽合肥230088

出　　处：《雷达科学与技术》2010年第5期463-468,共6页Radar Science and Technology

基　　金：国家863计划(No.2009AA12Z329)

摘　　要：提出了基于扩展的H∞滤波和地心坐标系的自适应多传感器误差配准方法,该方法采用地心坐标系来消除地球曲率对配准算法的影响,采用扩展的H∞滤波来消除状态模型和量测模型噪声方差的不准确对配准算法的影响,该方法能够解决多传感器异步数据的误差配准问题,并且状态方程和量测方程的噪声可以是非高斯的,并采用修正的基于BLUE的量测转换方法对修正后的传感器的量测进行滤波。仿真试验表明,该算法能有效地估计时变的传感器误差,即使在运动方程和量测方程的噪声方差不准确的情况下,也能得到满意的配准结果。In this paper,an adaptive multi-sensor registration algorithm based on extended H∞filtering and ECEF coordinate system is presented.The proposed algorithm applies ECEF coordinate system to re-move the influence of the earth curvature,and adopts extended H∞filtering to decrease the influence of inac-curate variance of the dynamic model and measurement model.The algorithm can resolve multi-sensor asyn-chronous data registration and the noises of the dynamic model and measurement model are not always Gauss-ian distribution.At last,the measurement data with correct systematic errors are applied to the filter by using modified converted measurement algorithm based on BLUE.The simulation results show the effective-ness of the registration algorithm.Even though the variances of the dynamic model and measurement model are not inaccurate,the algorithm also achieves satisfied results.

关 键 词：误差配准 扩展H∞滤波 ECEF坐标系 多传感器 

分 类 号：TN957[电子电信—信号与信息处理]