Amortized Global Search for Efficient Preliminary Trajectory Design with Deep Generative Models
This work addresses the high computational cost in trajectory optimization for aerospace engineering, though it appears incremental as it builds on existing deep generative methods for similar problems.
The paper tackles the computationally demanding global search problem in preliminary trajectory design by exploiting clustering in solutions and proposing an amortized global search framework using deep generative models to accelerate search for unseen parameters, achieving faster convergence in evaluations on De Jong's 5th function and a low-thrust circular restricted three-body problem.
Preliminary trajectory design is a global search problem that seeks multiple qualitatively different solutions to a trajectory optimization problem. Due to its high dimensionality and non-convexity, and the frequent adjustment of problem parameters, the global search becomes computationally demanding. In this paper, we exploit the clustering structure in the solutions and propose an amortized global search (AmorGS) framework. We use deep generative models to predict trajectory solutions that share similar structures with previously solved problems, which accelerates the global search for unseen parameter values. Our method is evaluated using De Jong's 5th function and a low-thrust circular restricted three-body problem.