利用支持向量机的矢量跟踪通道故障检测算法

为了克服卫星导航矢量跟踪接收机中故障通道对其他正常通道的影响,提高矢量跟踪接收机的鲁棒性,提出一种利用支持向量机的通道故障检测算法,利用矢量跟踪接收机的导航滤波器的新息序列作为支持向量机的输入。支持向量机的输出为通道的状态,根据支持向量机的输出通道的状态来判断该通道是否纳入导航滤波器,用来跟踪滤波器状态值,这样能够有效地避免故障通道对导航结果的污染。仿真实验结果表明:该方法能准确地检测出有故障的通道,提高矢量跟踪环路的鲁棒性。

Applying SVM to Fault Channel Detection in VectorTracking Loop

Vector tracking loop is one of the advanced technology of the satellite navigation receiver. The algorithm employs a central kalman filter to complete the tracking and navigation, and then calculate the loop parameter using the navigation solutions. Making full use of the shared information of the different channels can better the performance of the anti-interference and bridge the instantaneous blocked signals. However, the disadvantage is that the fault channel could influence other channels and cause the receiver fail to provide navigation information. This paper proposed a channel fault detection algorithm, which employs a trained SVM to classify the channel. The simulation results show that this method can accurately detect faulty channels, improve the performance and robustness of the vector tracking loop.