Abstract

Accurate and efficient modeling of complex dynamical systems remains a central challenge in modern astrodynamics, especially for emerging applications in cislunar space and the growing orbital debris problem. Dr. Tahsinul Haque Tasif’s research focuses on computational astrodynamics through the development of semi-analytical methods and the integration of machine learning techniques.

In this talk, he will present two applications of his research: high-precision trajectory propagation in multi-body dynamical systems and machine learning–based prediction of orbital debris evolution. He will first discuss an analytic continuation framework for propagating spacecraft states and state transition matrices in the Circular Restricted Three-Body Problem, enabling high-accuracy solutions with reduced computational cost and supporting applications such as efficient computation of Finite Time Lyapunov Exponent (FTLE) maps for dynamical systems analysis to support Earth–Moon trajectory design and navigation. He will then present a hybrid framework that combines physics-based debris breakup models, semi-analytical orbital propagation and deep neural networks to efficiently predict the evolution of debris populations in low Earth orbit. Finally, he will discuss future directions of his work involving higher-order state transition tensors, full ephemeris models and onboard machine learning for autonomous spacecraft operations.