【作者】 朱帅林； 李子涵； 唐子扬； 王振涛； 冯彦翔； 杨宜康；

【Author】 Shuailin Zhu;Zihan Li;Ziyang Tang;Zhentao Wang;Yanxiang Feng;Yikang Yang;School of Automation Science and Engineering, Xi’an Jiaotong University;

【机构】 西安交通大学自动化科学与工程学院；

【摘要】 随着空间技术的快速发展与卫星规模的迅猛增长,如何充分利用地面站资源,对大量测控任务进行高效调度正成为一个极具挑战性的问题。任务优先约束的存在限制了一组任务在执行顺序上的关系,为问题求解带来了新的挑战。为了解决该问题,本文提出一种考虑任务优先约束的离散Jaya算法(DJPC)。本文首先建立考虑任务优先约束的卫星测控资源调度模型,然后以趋近最优解和规避最劣解为目标建立离散Jaya算子作为DJPC的核心进化机制。为了减少时间窗冲突导致的收益损失,本文引入一种基于时间窗冲突风险的启发式搜索策略,该策略有助于提高搜索效率,并避免算法陷入局部最优。针对任务优先约束,提出一种基于前界插入和后界插入的约束修复策略,确保算法的求解结果满足任务优先约束。与其他算法的对比实验证明,DJPC的求解结果对于较大规模的调度场景(待调度任务数不少于500)具有显著优势。更多还原

【Abstract】 With the rapid development of space technology and the rapid increase in the number of satellites, how to make full use of ground resources and efficiently schedule a large number of tasks is becoming a very challenging problem. The existence of task precedence constraints restricts the relationship of a group of tasks in the execution sequence, bringing new challenges. To solve the problem, a discrete Jaya algorithm considering task precedence constraints(DJPC) is proposed. First, establish a satellite range scheduling model considering task precedence constraints, and then propose a discrete Jaya operator as the core evolutionary mechanism of DJPC. In order to reduce the profit loss caused by time window conflicts, a heuristic search strategy based on the risk of time window conflicts is introduced, which helps to improve search efficiency and avoid falling into local optima. In addition, we propose a constraint repair strategy based on the method of forward-bound insertion and backward-bound insertion to ensure that the solutions meet the precedence constraints. Finally, the effectiveness of each part of the DJPC is verified through experiments. Comparative experiments with other algorithms prove that the solution of DJPC have an significant advantage in large-scale scheduling scenarios(with no fewer than 500 tasks).更多还原