月球低能返回轨道设计的混合自适应遗传算法

An adaptive genetic algorithm for low energy lunar return trajectory design

摘要:

针对月球探测器低能量返回轨道设计问题，在椭圆四体问题下建立探测器动力学模型，考虑太阳引力与月球椭圆运动对探测器轨道的影响，并分析探测器低能量返回轨道的存在性与轨道动力学特性，以及针对设计月球最低能量返回轨道过程中模型强非线性、全局优化性能差等现象，提出了一种混合自适应遗传算法用于寻找月球探测器返回地球所需的最低能量以及对应的转移轨道，算法根据种群适应度数据自适应改变进化特征，提高了种群向全局最优点进化的效率，降低了计算量.仿真结果表明，此种算法可以精确高效地计算月球探测器返回地球所需的最低能量，所需的速度脉冲仅为传统的双曲拼接法的75%，显著节约了能源消耗.

Abstract:

To design low energy return trajectory for unmanned lunar probe, the dynamic model of the probe is developed under the elliptical four body problem in consideration of the effect of sun's gravitation and lunar elliptical motion to the probe's orbit. The existence and the dynamical characteristics of the return trajectory are analyzed. To deal with the strong nonlinearities of the dynamic model and the local convergence in the optimal process, an adaptive genetic algorithm is proposed which can adjust the self-evolution parameters according to the fitness of the population to improve the effectiveness of the evolution of the population to the global optimal point as well as to reduce the computation burden. According to the simulation results, the algorithm works well in optimizing the energy needed for the lunar probe to return to Earth, which is only 75% of that in the traditional hyperbolic matching method.