INTELLIGENT AIRCRAFT TRAJECTORY PREDICTION WITH AN OBJECTIVEDRIVEN DIFFUSION MODEL

Abstract

Aircraft trajectory prediction plays a critical role in modern air traffic management systems, enabling safer navigation, efficient route planning, and improved airspace utilization. Traditional trajectory prediction approaches often rely on statistical models or deterministic algorithms that struggle to capture the complex and dynamic nature of flight movements influenced by weather conditions, air traffic constraints, and operational objectives. To address these challenges, this study proposes an Intelligent Aircraft Trajectory Prediction framework using an ObjectiveDriven Diffusion Model. The proposed approach leverages diffusion-based deep generative models to learn complex spatiotemporal patterns from historical flight trajectory data. By incorporating goal-oriented objectives such as destination points, flight constraints, and operational priorities, the model can generate accurate and realistic future flight paths. The diffusion process gradually refines noisy trajectory representations into precise predictions, enabling the system to capture uncertainty and variability in aircraft movements. Furthermore, the framework integrates contextual information including flight dynamics, environmental factors, and airspace regulations to enhance prediction reliability. The objective-driven mechanism guides the model toward producing trajectories that satisfy operational goals while maintaining safety and efficiency. Experimental evaluations demonstrate that the proposed diffusion-based framework significantly improves prediction accuracy and robustness compared to traditional machine learning and sequence-based models. The system supports advanced air traffic management applications such as collision avoidance, route optimization, and real-time flight monitoring, contributing to safer and more efficient aviation operations.