渐消SPRT在SINS/CNS/SRS导航系统软故障诊断中的应用

Fading SPRT method for soft fault diagnosis in SINS/CNS/SRS integrated navigation system

为提高导航系统的可靠性,实时的故障诊断与隔离十分必要。序贯概率比检验(SPRT)对传感器中缓慢增长的软故障具有较高的灵敏度,但该方法却存在故障检测延迟甚至无法检测故障结束的缺陷,造成故障的漏警和误警。对SPRT在故障检测中存在的缺陷进行了分析并提出了一种渐消SPRT方法。渐消SPRT方法通过引入渐消因子,降低了历史信息对故障时刻统计量变化率与统计量的影响,进一步增强SPRT对故障的灵敏度,从而实现降低故障检测延迟并避免SPRT无法检测故障结束的问题。最后基于捷联惯导/天文/光谱红移(SINS/CNS/SRS)组合导航系统,进行了导航系统实施故障诊断仿真验证。仿真结果表明相比于SPRT,渐消SPRT能够检测到缓变故障的结束并将故障开始时刻的检测延迟降低了41%,极大地提高了SINS/CNS/SRS组合导航系统的实时估计精度与可靠性。在实时导航系统中,渐消SPRT能够起到良好的软故障检测与隔离作用,保证了系统的稳定性。

In order to improve the stability of the navigation system,real-time fault diagnosis and isolation are essential.The sequential probability ratio test(SPRT)has sensitivity to the detection of slow-growing soft faults,but SPRT has an obviously defect that has time delay in fault detection and even cannot detect the fault end.The defect of SPRT is analyzed and further a fading SPRT method is proposed.The Fading SPRT method introduces the Fading factor in the calculation of SPRT statistics,which reduces the influence of historical measurement on the statistics at the time with failure,thus avoids the problem that SPRT cannot detect the end of failure and reduce the delay in fault detection.Finally,based on the proposed SINS/CNS/SRS integrated navigation system,the fading SPRT method is verified.As the simulation results shown,fading SPRT can detect the fault end and reduced the time delay in fault start detection by 41%,which greatly improves the real-time estimation accuracy and reliability of SINS/CNS/SRS integrated navigation system under slow-growing fault.