An Adaptive Time-Series Probabilistic Framework for 4-D Trajectory Conformance Monitoring

Trajectory conformance monitoring is important for future air traffic control for reasons associated with optimal operation, increased safety, and improved efficiency. In this study, conformance monitoring is considered with respect to preassigned 4-D (space and time) trajectories and their margins (4-D contracts), and an adaptive time-series probabilistic framework is postulated. Two problems are tackled, and proper methods are developed: 1) present conformance monitoring and quality of conformance evaluation via statistical tools, which leads to abnormal event detection; and 2) future conformance monitoring in which conformance is predicted ahead of time allowing for the early initiation of corrective actions. The framework is based on recursive integrated autoregressive modeling of contract deviations alone, with the underlying dynamics and nonstationarity accounted for. An initial assessment of the performance of the framework is based on two simulation scenarios. Through them, present conformance monitoring is shown to lead to quality assessment and the declaration of an alarm immediately following the emergence of an abnormal event. Future conformance monitoring is shown to lead to an early nonconformance alarm, with the lead time shown to be significantly longer than that achieved by a current probabilistic benchmark scheme.