应用天文研究组知识库

结合当今技术的发展状况，针对台站观测任务数量多、问题建模难、求解复杂度高等问题，提出了一种基于贪婪算法与动态规划算法融合的改进贪婪动态规划算法。该算法首先将调度问题划分成几个子问题，然后根据约束条件以最大化观测收益为目标函数对问题进行迭代求解，进而生成近似最优观测计划。实验结果表明，该算法在解决观测任务调度问题时具有一定的可行性和实用性，并为下一步建立台站自动化运行系统奠定坚实的基础。 

Combining with the development of current technology, this paper proposes an improved greedy dynamic programming algorithm based on the fusion of greedy algorithm and dynamic programming algorithm for the problems of large number of station observation tasks, difficult problem modeling, and high solution complexity. First, this algorithm divides the scheduling problem into several subproblems, iteratively solves them with the objective function of maximizing the observation income according to constraints and generates an approximate optimal observation plan. Experimental results show that this algorithm has certain feasibility and practicability in solving observation task scheduling problems, and lays a solid foundation for the next step of establishing an automated station operation system.