Abstract: In order to find the best on/off sequence of thruster in multi pulse guidance for rendezvous of spacecraft to realize propellant consumption optimization, a genetic algorithm based on model the predictive and inversive control is proposed. In this method, the fitness function of genetic individual is designed according to total speed increment needed for orbit control. The genetic chromosome is constructed in binary coding form, each gene position is set to be 0 or 1, corresponding to whether the thruster works in a control cycle. The heuristic strategy of equidistant sparse control is used to generate the initial population, and the penalty functions are used to satisfy the engineering constraints. By iterative operation of genetic crossover, mutation and selection, the propellant consumption of multi pulse rendezvous is optimized. Simulation results show that the on/off sequence of thruster optimized by the genetic algorithm has obvious characteristics, which can effectively shorten the working time of thruster in guidance and reduce the speed increment needed for orbit control, thus reducing the propellant consumption and still ensuring the guidance accuracy.

Key words: Rendezvous and docking, Multi pulse guidance, Model predictive and inversive control (MPIC), Propellant consumption, Genetic algorithm, Trajectory optimization