A Light Statistical Method of Air Traffic Delays Prediction

Abstract

An air traffic delay is a common problem of a modern transport system. Continuously increasing demand of air traffic system and operation in a rapidly changing environment makes delay estimation is a key element in aviation safety. In the paper, we propose a light statistical method of airplane landing delay prediction at the destination airport. Proposed method grounds on statistical data processing of previous flights to identify mean trajectory of an airplane at the pre-flight stage and continuously comparison of mean trajectory with current airplane location in order to identify a time to land in the destination airport. A simple solution of minimization problem at the en-route phase of flight needs low computation power and guarantees the required confidence band for results. Proposed method has been trained on statistical data for Boryspil / Kharkiv connection routes in order to airplane delay prediction.