A new approach to aircraft categorization using machine learning to analyze aircraft behaviour

Abstract

The establishment of aircraft categories is a classification technique employed in a variety of aviation disciplines including design and development, certification, ongoing airworthiness, air traffic management, surveillance, and safety analysis. Traditional approaches to aircraft classification rely on manual feature engineering which can be labor-intensive and ineffective for capturing complex patterns. In this paper, a novel approach to aircraft classification using unsupervised machine learning clustering techniques is proposed. The aim of the proposed approach is to be simple in order to be useful and understandable across disciplinary domains; to be scalable to large volumes of air traffic data in order to leverage this data for the purpose of improving the understanding of aircraft behaviours; and to be easily adaptable to future changes in order to account for the evolving technological and operational nature of the airspace environment. The application is based on an adapted version of the k-means algorithm that can automatically group aircraft into clusters based on similarities in features such as position, velocity and acceleration over a period of time. The proposed approach is validated using a real-world air traffic dataset obtained from publicly available ADS-B data, and the results are compared to traditional classification methods from the field of aircraft certification. The findings suggest that leveraging machine learning clustering techniques is a promising approach to aircraft classification, enabling automated and scalable solutions for a variety of applications. The proposed approach has the potential to improve aircraft classification accuracy tailored for specific applications, enhance situational awareness of the airspace environment, and ultimately enhance aviation safety.