摘要
理想情况下,基于卡尔曼滤波的组合导航系统在无故障时的滤波残差为白噪声,故障时通常用2χ法来检测,但用这种方法检测往往具有较大的虚警。因此,针对无故障时卡尔曼滤波器的滤波残差是白噪声这一特点,提出了一种新的故障检测方法来判断系统是否发生了故障,在应用时大大降低了虚警,取得了较好的效果。
Ideally,filter residuals is a white noise in the Kalman filter-based integrated navigation system in the absence of fault,and χ2 method usually falls to detect while failure ocuurs.But detection in this way often brings higher false alarm.Therefore,the feature for Kalman filter residual is white noise when no fault,this paper proposed a new fault detection method to determine whether the system malfunctioned,and it′s performed well when applicated.
出处
《航空计算技术》
2011年第2期132-134,共3页
Aeronautical Computing Technique
基金
军队科研基金项目资助
关键词
卡尔曼滤波
故障检测
遍历性
白噪声
kalman filter
fault detection
ergodicity
white noise
