Interacting Multiple Model Filtering for Aircraft Guidance Modes Identification from Surveillance Data

Abstract

Accurate and reliable trajectory prediction (TP) is required in several air traffic management systems. Estimating the aircraft trajectory in a vertical plane typically requires the knowledge of a pair of operational instructions. A sequence of operational instructions specifies the aircraft intent, information that is seldom available, besides for the ownship trajectory planning system. In the execution of the trajectory, the aircraft is directed by the (auto)pilot through a series of sequential guidance modes that might override some of the planning instructions of the aircraft intent. Therefore, having reliable guidance mode information is fundamental for the next generation of air- or ground-based TP. The main goal of this contribution is to develop a methodology able to identify in real-time the active guidance modes for both vertical climb and descent profiles, using only Automatic Dependent Surveillance-Broadcast and Enhanced Mode-S Surveillance data. The proposed solution is based on an interacting multiple model (IMM) filtering approach, which uses a bank of filters, each one matched to a possible guidance mode. The guidance mode identification performance of the IMM-based solution is validated with i) a set of simulated representative trajectories and ii) real flight data obtained from flight data recorders.