TREATMENT OF FAULTED NAVIGATION SENSOR ERROR WHEN ASSESSING RISK OF UNSAFE LANDING FOR CAT IIIB LAAS

摘要

Of interest in the assessment of LAAS for CAT IIIb autoland operations is the risk of an unsafe landing due to a fault in the navigation sensor error (NSE). This paper presents two methods for grouping the three error components that comprise the total system error (TSE). These components are: fault-free NSE component (NSE_(ff)), faulted-NSE component (E) and flight technical error (FTE). The first of the two methods, known herein as the faulted-NSE method, treats the faulted component E as distinct and groups NSE_(ff) with FTE. The second method, known herein as the total-NSE method groups NSE_(ff) with E and therefore makes the distinction between the total NSE and FTE. The paper shows that the two methods are equivalent by a mathematical proof and an example application using a generic fault monitor. However, the paper goes on to show that the faulted-NSE method has an advantage over the total-NSE method if the distribution of errors due to a fault is not known. In that circumstance, the faulted-NSE method allows identification of a worst value of E, for which the conditional risk of unsafe landing is highest. However, the total-NSE method does not permit identification of a corresponding worst value of NSE. Therefore, the faulted-NSE method is used to address an existing limit case requirement for conditional risk of unsafe landing when the NSE contains a large faulted component. A method of specifying ground and airborne requirements for the limit case is chosen based on previously proposed alternatives. Based on representative aircraft landing performance characteristics, a requirement is then derived for the ground monitor missed detection performance. Compliance with this monitoring requirement is illustrated for a generic ranging source fault monitor.