Applying hybrid genetic algorithm to constrained trajectory optimization

Trajectory optimization is a typical optimal control problem. Aiming at the slow convergence characteristics and the poor local searching ability of a basic genetic algorithm, this paper proposed a new hybrid global-local optimization algorithm by coming genetic algorithm and complex algorithm to improve the convergence rate of genetic algorithm. The hybrid way adopted serial hybrid pattern in this paper. It can mean that the optimal solution of genetic algorithm is submitted as an initial parameter set to complex algorithm for refinement. In order to validate algorithm, hybrid genetic algorithm applied to lunar soft landing trajectory optimization problem. Simulation results demonstrate that the methodology and algorithms take on fast convergence rate and high optimization precision.