Low-thrust orbital transfers are difficult to optimize by indirect methods. The main issues come from the costate guess and from the numerical propagation accuracy required by the shooting method. In the case of a coplanar minimum-time low-thrust transfer with eclipses, an analytical costate guess is proposed. The optimal control problem reduces to an unconstrained minimization problem with two unknowns. A derivative-free algorithm yields a quasi-optimal solution from scratch in a few minutes. No specific guess is necessary and the algorithm properties ensure finding the global minimum of the unconstrained problem. The method is applicable to thrust levels from large to very low and to any eclipse configuration, as exemplified on a transfer towards the geostationary orbit.
Global Search for Optimal Low Thrust Spacecraft Trajectories using Diffusion Models and the Indirect Method
PhD Student, Department of Mechanical and Aerospace Engineering, Princeton University, NJ, USA（普林斯顿大学机械与航空航天工程系博士生）；Assistant Professor, Department of Mechanical and Aerospace Engineering, Princeton University, NJ, USA（普林斯顿大学机械与航空航天工程系助理教授）
Self-supervised diffusion model fine-tuning for costate initialization using Markov chain Monte Carlo
PhD Student, Department of Mechanical and Aerospace Engineering, Princeton University, NJ, USA（普林斯顿大学机械与航空航天工程系博士生）；Assistant Professor, Department of Mechanical and Aerospace Engineering, Princeton University, NJ, USA（普林斯顿大学机械与航空航天工程系助理教授）
