Abstract: |
Multi-objective genetic algorithms (MOGA) are used to optimize a low-thrust spacecraft control law for orbit transfers around a central body. A Lyapunov feedback control law called the Q-law is used to create a feasible orbit transfer. Then, the parameters in the Q-law are optimized with MOGAs. The optimization goal is to minimize both the flight time and the consumed propellant mass of the trajectory created by the Q-law, and consequently to find Pareto-optimal trajectories. To improve the qualities of the obtained Pareto-optimal solutions, elitism and a diversity-preservation mechanism are incorporated into MOGA. The MOGA performance with and without the new mechanisms are systematically compared and evaluated with quantitative metrics. The new mechanisms significantly improve the convergence and distribution of the resulting Pareto front for the low-thrust orbit-transfer optimization problem. The new mechanisms also improve the statistical stability and the computational efficiency of the algorithm performance. |
