抗野值强跟踪Kalman滤波在风场估计中的应用  

Application of strong tracking Kalman filtering with restraining outliers in wind field estimation

作　　者：李阳[1,2] 孔毅[1] 赵现斌[1] 

机构地区：[1]解放军理工大学气象海洋学院 [2]中国人民解放军94994部队

出　　处：《传感器与微系统》2015年第7期147-150,共4页Transducer and Microsystem Technologies

基　　金：国家自然科学基金资助项目(41076118)

摘　　要：由于受周围环境干扰和传感器短暂性失效等因素的影响,无人机风场测量数据中包含较大野值或成片野值,影响测量数据的准确性。根据无人机测风的特点,结合Kalman滤波算法、强跟踪滤波算法和抗野值修正算法的优点,通过对滤波发散趋势的分析,提出上述滤波算法的使用条件,构造了一种抗野值抑制发散滤波算法,并进行了实验论证。实验结果表明:该算法能有效克服野值对滤波造成的不良影响,具有良好的抗野值能力、跟踪能力,保证了滤波精度,可适用于无人机风场测量。Due to influence of ambient interference and transient sensor failure and so on, wind field measurement data with UAV contains the larger or more continuous outliers which affect accuracy of measurement data. According to characters of wind field measurement by UAV and combine with merits of Kalman filtering algorithm,strong tracking filtering algorithm and restraining outliers correcting algorithm,through analysis on filtering divergence trend,using conditions of above filtering algorithms are proposed,eventually a filtering algorithm with restraining outliers and divergence is constructed and test verification is carried out. Test result indicates that the algorithm can effectively conquer the influence of outliers,has good ability of restraining outliers,tracking ability and ensure filtering precision,and it can be applied for wind field measurement by UAV.

关 键 词：无人机 KALMAN滤波 强跟踪Kalman滤波 野值 新息 

分 类 号：P414.2[天文地球—大气科学及气象学]