The optimal ascent trajectory design of the vehicle is an optimal control problem with strict terminal constraints and variable final time. The classical algorithms always encounter the problems of high sensitivity to initial guess and local convergence in solving this problem. Aiming at these problems, genetic algorithm (GA) which is of good global convergence is applied to designing the optimal ascent trajectory .In order to improve the convergence rate of GA and overcome its premature problems, two hybrid GA (HGA) combining the merits of GA with those of simplex method are proposed. The computational results testify that the two HGA are effective global optimization approach for solving complex problems, and they can successfully solve the optimal control problems with variable final time.