基于多子人口群协作进化的拦截轨道优化

Optimization of Intercept Orbit Based on Multi-Subpopulation Cooperation Evolution

针对基本遗传算法对航天器拦截轨道数值优化计算效率较低的问题,提出了一种新的基于多子人口群协作进化的算法。使用子人口划分技术提高了人口多样性防止早熟,用免疫算子减小搜索空间,两者都加速了进化计算过程。应用此算法求解了具有推力约束和拦截时间约束使燃料消耗量最小的航天器拦截轨道,并分析了其与基本遗传算法的不同。通过航天器拦截轨道仿真表明,该算法优于基本遗传算法,可用较少的计算时间得到全局最佳解,提高了航天器拦截轨道优化的计算效率。

To overcome the difficulty that the numerical optimization of intercept obit is too time consuming using the basic genetic algorithm, this paper considers the numerical optimization problem of intercept obit based on evolution computation. A novel multi-subpopulation cooperation genetic algorithm is put forward and analyzed here in detail, in which the enforced subpopulation technique improves the diversity of the population to avoid premature stop and immunization technique decreases the searching space and they both accelerate the search to reach the global minimum. By this method, a minimum fuel interception-orbit problem with finite-thrust and bounded-time is successfully solved by the multi-subpopulation cooperation evolution. The differences between these two methods are also analyzed. The simulation data of the interception orbit prove that the new method is better than the basic genetic method in that less computation time is required for reaching global minimum, the numerical optimization of intercept obit can be completed with high efficiency.