多敏捷卫星协同任务规划调度方法

摘要

The agile satellite mission planning is a multi-target combination with a long window,multiple time windows and other complex optimization problem.Based on the image quality,aimed to build an attitude agile earth observation satellite mission planning Portfolio Optimization Model through analysis of attitude agile satellite needs,characteristics and constraints,on the basis of the original simulated annealing algorithm,a genetic simulated annealing hybrid algorithm was designed.By the similarity degree and aggregation degree,the mutation probability of the chromosome was increased when the population aggregation degree was large,so as to increase the population diversity.The global searching ability of genetic algorithm is favorable to change the drawback that simulated annealing algorithm is easy to fall into the local minimum point and find better results,which makes the algorithm achieve the balance of global searching ability and local searching ability.The algorithm is validated by the actual satellite mission data.%敏捷卫星任务规划调度是一个具有长时间窗、多时间窗的复杂约束的多目标组合优化问题.基于任务质量,通过分析敏捷卫星对地观测任务规划问题的需求、特点和约束,构建了敏捷卫星任务规划组合优化模型;并在原有模拟退火算法的基础上,设计了基于相似度和聚集度的遗传模拟退火混合算法,通过相似度和聚集度,在染色体变异过程中,当种群聚集度大的时候,增加染色体的变异概率,从而增加种群的多样性.利用遗传算法的全局搜索能力有利于改变模拟退火算法容易陷入局部最小点的缺点,寻找到更优的结果,使算法达到全局搜索能力与局部搜索能力的平衡,经实际卫星任务数据验证算法有效可行.