基于伪距误差重建的多星故障检测方法

Multiple failure detection based on reconstruction of the pseudorange error

传统的接收机自主完好性监测(RAIM)方法使用定位解算方程的残差来进行故障识别,在单卫星故障下具有较好的性能。随着全球卫星导航系统的建设,用户观测到的卫星数目显著增加,使用多颗卫星定位可以提高定位精度。但多颗卫星同时发生故障的概率将会增大,RAIM方法需要进行改进以应对多星故障。该文首先对RAIM方法的原理进行了深入分析,推导出了误差向量和残差向量之间的投影关系。然后,通过实例论证了误差向量和残差向量在投影变换时会带来信息损失,从而使故障难以检测。最终,给出了一种新的RAIM方法。通过引入约束条件,利用残差向量重构出误差向量。根据误差向量大小,实现多星故障检测。仿真表明该方法在多星故障模式下,具有较好的检测性能。

The traditional RAIM algorithm uses the residuals of the positioning equation for failure detection and gives better detection when only one fault exists. The global navigation satellite system gives users many more ranging sources with better positioning thanks to the increased number of visible satellites. However, the probability of simultaneous failures also increase which requires improves to the traditional RAIM algorithm to address multiple failures. The RAIM design was analyzed to deduce a relationship between the residuals and the error vector. Examples then show that some information is lost when projecting from the error vector to the residuals, which limits failure detection. A modified RAIM algorithm with additional constraints that enable the error vector to be recovered from the residuals so that multiple failures can be detected monitoring the error vector. Simulations show the ability of this method for multiple failures.