针对目前GPS接收机自主完好性监测方法受到许多假设的制约这一问题,该文根据已有的导航系统可靠性研究,基于非理想故障模式以及故障分布参数的不确定性,利用先验概率模型对卫星和接收机异常建模;采用该模型,建立了基于误警率及漏检率的最小风险代价任意目标函数,以对残差统计量分析进而优化选取阈值;最后对快照接收机自主完好性监测法以及利用贝叶斯更新得到后验故障概率的多步完好性监测算法进行分析,具有一定的参考意义,多步算法将会进一步丰富GPS完好性结构。仿真实验结果表明,在每个几何条件下用反复试验法可进行阈值的搜索,达到比传统接收机自主完好性监测方法更好的效果。
Current methods for GPS receiver autonomous integrity monitoring would be limited by many assumptions which must be taken.According to the existing achievement on navigation system reliability,based on non-ideal failure models and the uncertainty of failure distribution parameters,this paper presented a prior probability model for satellites and receiver anomalies.Using this model,the minimal cost function were founded based on the false alarm and missed detection errors to analyze the residuals test statistic and select the optimal threshold.At last,the paper analyzed the traditional snapshot receiver autonomous integrity monitoring(RAIM)methods as well as the multiple integrity algorithms which were based on the posterior failure probabilities generated by Bayesian updating.The simulation experiment results indicated that this method could get better performance than traditional methods.