摘要
通过分析广义似然比方法,指出该方法不能用于多星故障识别。为能够快速识别多颗故障卫星,提高卫星导航定位系统的可靠性,提出了一种可以用于多颗故障卫星识别的接收机自主完整性监测(RAIM)的新方法——假设验证法。该方法对所有可能的故障卫星组合进行假设验证,对每种假设的故障卫星组合计算出卫星伪距偏差,然后利用此偏差构造一个新的奇偶残差矢量,最后利用特定的故障识别准则进行判断。以识别2颗故障卫星和识别3颗故障卫星为例进行了计算机仿真,结果显示:假设验证法故障正确识别率高于85%,高于现有的可用于多星故障识别的最优奇偶矢量法,可以有效提高卫星导航系统的可靠性。同时,与最优奇偶矢量法相比,假设验证法不需要求取矩阵广义特征向量,计算量将减少90%以上。
The generalized likelihood test method is studied, and the result shows that it is unable to identify the simultaneous multi-faulty satellites. In order to identify multi-faulty satellites rapidly, and to improve the reliability of satellites navigation system, a new receiver autonomous integrity monitoring（RAIM） algorithm is proposed, which is called assuming and validating method. This method validates every possible faulty satellites combination, computes satellite pseudo-range deviation for each combination, and then eonstruetes a new parity residual vector by the deviation, finally provides the judging result with the judging function. Computer simulation is achieved under the conditions of simultaneous two-faulty satellites and simultaneous three-faulty satellites, and the results indicate that compared with the other existing methods of the same kind, the assuming and validating method increases correct identifying ratio which will be higher than 85%. This method will improve the reliability of satellites navigation system. Moreover, compared with the optimal parity vector method, this method avoids computing the generalized eigenvectors of matrix, and its computation load decreases more than 90 %.
出处
《航空学报》
EI
CAS
CSCD
北大核心
2008年第5期1239-1244,共6页
Acta Aeronautica et Astronautica Sinica
基金
国防科工委民用航天重大专项
