MODELING AIRPLANE SURVEILLANCE SYSTEM TO ENHANCE SECURITY USING MACHINE BASED ALGORITHM

Abstract

Effective air traffic management is crucial for ensuring safety, reducing delays, and optimizing operational efficiency in aviation. This study presents the development of an Airplane Trajectory Prediction (ATP) model for Air Traffic Management System (ATMS) using a Support Vector Machine (SVM) algorithm. The aim of the study is to provide real-time trajectory monitoring and anomaly detection while classifying flight behaviours into three categories: normal flight, in-flight delay, and high-speed. A secondary dataset of aircraft trajectories was pre-processed using methods including mean imputation and Z-normalization as part of the approach. Flight behaviours were analysed using a multi-class SVM classifier, and hyperparameter adjustment was done to get the best prediction accuracy. The result showed how well the model performed in correctly predicting and categorizing flight paths. The ATP model provided substantial advantages for operational and safety decision-making by successfully detecting possible abnormalities and diverging from intended routes with 89% success rate. To further increase prediction accuracy and scalability, future studies might investigate the use of more advanced deep learning models and bigger datasets.

KEYWORDS:  Airplane Trajectory Prediction (ATP); Air Traffic Management System (ATMS); Support Vector   Machine (SVM); Machine Learning; Z-Normalization; Aircraft Trajectories