An Automatic Flight Go-around Maneuvers Identification Method Based on Action Feature Analysis

“Go-around” is an abnormal status that requires significant attention in the operational control and support business of airlines, as it directly affects the safety and efficiency of flight operations. Thanks to the popularization of technologies, air traffic control and airlines have basically achieved full-time monitoring of flight operations. However, the perception of abnormal flight statuses still lacks timely and automatic identification methods, often relying on the proactive discovery of operational control personnel or post-hoc reports from the flight crew, which is highly disadvantageous for handling abnormal flight conditions. To address this issue, a method for identifying go-around maneuvers based on real-time ADS-B data is proposed in this article. This method utilizes real-time aircraft trajectory data, combined with information such as flight procedures and runways, to construct go-around maneuvers sensing areas and models. It then forms a vector-based structural model for go-around maneuvers identification, enabling the determination of aircraft stable approach status, automated identification of go-around maneuvers, and automatic classification and alerting of go-around maneuvers. Considering the varying risk levels of go-arounds under different conditions, the go-around maneuvers are scientifically classified based on the status information at the time of the go-around, thereby improving operational control efficiency. The impact of changes in runway operation directions on identification results is excluded, and the judgment of runway operation directions is also taken into account. Experiments show that this method has high identification accuracy and low dependency costs, effectively solving the problem of real-time perception of go-around maneuvers and fully satisfying the business needs of flight operations support.