Optimal Fuel Solution to the Lambert Problem under Complex Perturbation Conditions Based on BP Neural Network

Addressing the orbital interception problem, various perturbations are considered initially, and the impact of perturbations on solutions in the two-body problem is analyzed. Numerical analysis is then introduced to successfully mitigate the effects of perturbations. Finally, inspired by the concept of BP neural networks, an original training set is constructed based on the numerical correction of perturbed Lambert problem solutions. This training set is used to train a neural network to approximate solutions to the perturbed Lambert problem. The resulting neural network exhibits performance similar to the numerical correction method, saving time and reducing computational costs while ensuring control accuracy.