在軌衛星的智能規劃可以歸結為資源受限的實時規劃和調度問題，由于此類問題的規
模大，實時性要求強，通常的解法很難滿足應用要求。本文針對該問題的特點，建立了問題的數學模型，包括知識表示、目標函數等。提出了三種啟發式的規劃算法，并對算法特點進行了比較分析，具體應用時，可以根據問題的不同特點選用相應的算法，從快速性和準確性兩方面適應系統要求。最后本文通過一個算例，從統計特性角度分析了三種算法，驗證了算法的正確性。
關鍵詞: 智能規劃；啟發式算法；資源約束；實時性
【Abstract】The intelligence scheduling and planning for on-board satellite belongs to scheduling problem under resources constrains. The problem is about large amount of tasks, and have tight real-time deadlines, so the common algorithm is hard to satisfy the requirement. This paper presents three kinds of heuristic algorithms, and make the fully analysis. In application, we can choose the proper algorithm base on the feature of the problem. Finally, Simulation results show the feasibility of the algorithm.
【Key words】intelligence scheduling; heuristic algorithm； resource-constrained； real_time