Cost-efficient scheduling of FAST observations
Cost-efficient scheduling of FAST observations
复制标题
经济高效地安排 FAST 观测
DOI：
10.1007/s10686-018-9577-2
复制
发表时间：
2018
影响因子：
3
通讯作者：
Zhao LP
中科院分区：
文献类型：
--
作者：
Luo Qi;Yu Ce;Sun Jizhou;Zhao Laiping;Xiao Jian;Zhu Ming;Zhong Yi;Zhao LP
500米口径球面射电望远镜（FAST）的成本效益计划要求最大化可观测方案的数量和总体科学优先级，最小化望远镜移位产生的总体回转成本，同时考虑天体可见性、用户定义的可观测时间、避免射频干扰（RFI）等约束条件。在这方面的贡献，首先，我们解决的问题，最大限度地提高可观察的建议和科学的优先级建模为最小成本最大流（MCMF）的问题。最优调度可以通过任何MCMF求解算法找到。然后，为了最小化生成的调度的回转成本，我们设计了一个最大匹配的边缘检测为基础的方法，以减少问题的大小，并提出了一个回溯算法，以找到最小回转成本的完美匹配。在NASA/IPAC河外星系数据库（NED）的真实的数据集上的实验表明，该调度器可以在很短的时间内提高高科学优先级的可用时间的利用率，并显著降低转换开销.
A cost-efficient schedule for the Five-hundred-meter Aperture Spherical radio Telescope (FAST) requires to maximize the number of observable proposals and the overall scientific priority, and minimize the overall slew-cost generated by telescope shifting, while taking into account the constraints including the astronomical objects visibility, user-defined observable times, avoiding Radio Frequency Interference (RFI). In this contribution, first we solve the problem of maximizing the number of observable proposals and scientific priority by modeling it as a Minimum Cost Maximum Flow (MCMF) problem. The optimal schedule can be found by any MCMF solution algorithm. Then, for minimizing the slew-cost of the generated schedule, we devise a maximally-matchable edges detection-based method to reduce the problem size, and propose a backtracking algorithm to find the perfect matching with minimum slew-cost. Experiments on a real dataset from NASA/IPAC Extragalactic Database (NED) show that, the proposed scheduler can increase the usage of available times with high scientific priority and reduce the slew-cost significantly in a very short time.