一种用于行星间转移轨迹优化的混合优化方法

A Hybrid Optimization Method for Interplanetary Transition Trajectory Optimization

本文针对行星间转移轨迹优化问题,提出一种结合Gauss伪谱法和遗传算法的混合优化方法。这种混合优化方法通过遗传算法全局寻优获得Gauss伪谱法中所用的状态变量及控制变量初值,然后Gauss伪谱法根据遗传算法获得的初值进一步寻优,从而克服了传统间接优化法和直接优化法对初值的依赖。本文以地球至火星低推进过渡轨迹优化问题为例,对所提出的混合优化方法进行验证。仿真结果表明,这种优化方法既具有遗传算法的全局寻优能力又具有Gauss伪谱法的局部强收敛特性,能够在无任何初值猜测的情况下完成对近似全局最优解的搜索。

In this paper, a hybrid genetic algorithm Gauss pseudospectral method is proposed for the inter-planetary transition trajectory optimization. The initial guesses for the state and control histories used in the Gauss pseudospectral method are interpolated from the best solution of a genetic algorithm. The minimum-time low-thrust transfer trajectory from a geocentric circular orbit to an areocentric circular orbit was computed to verify the proposed hybrid optimization method. The numerical results show that the proposed hybrid approach effectively includes global search ability of genetic algorithm and the high convergence rate of sequential quadratic programming, and has the capability to search the feasible and global optimal solution without any initial value guess.