Skip to Main Content
This paper presents the application of a neuro-fuzzy learning approach to classify air traffic control (ATC) trajectory segments from recorded opportunity traffic. This method learns a fuzzy system using neural-network theory to determine its parameters (fuzzy sets and fuzzy rules) by processing data samples. The problem is prepared for analysing the Markov-chain probabilities estimated by an interacting multiple model (IMM) tracking filter operating forward and backward over available data. The performance of this data-driven classification system is compared with a more conventional approach based on transition detection on simulated and real data of representative situations. The problem's formulation for this application enabled an accurate classification of manoeuvring segments and the derivation of rules that explain the relation between input attributes and motion categories used to describe the recorded data.