ADMISSIBLE N-IMPULSE ORBIT TRANSFER AND RENDEZVOUS SOLVED USING A LEARNING OPTIMIZATION ALGORITHM

The N-impulse orbit transfer and orbit rendezvous problems have been constructed for use with evolutionary algorithms (EAs), where the total v of the maneuver is minimized constrained by a maximum allowed time of flight and maximum allowable v per impulse. New, novel constraints on the v have been derived to make the N-impulse problem more computationally tractable, and then the construct for using EAs is developed. In this paper, a specific optimization technique, called the Learning Approach to Sampling Optimization, is tested and compared to the genetic algorithm (GA).

ADMISSIBLE N-IMPULSE ORBIT TRANSFER AND RENDEZVOUS SOLVED USING A LEARNING OPTIMIZATION ALGORITHM Conference Paper