节点文献

卫星—地面站数传调度模型及算法研究

Research on Satellite-Ground Station Data Transmission Scheduling Models and Algorithms

【作者】 李云峰

【导师】 武小悦;

【作者基本信息】 国防科学技术大学 , 控制科学与工程, 2008, 博士

【摘要】 卫星数传调度问题是指在地面资源有限的情况下,如何为卫星数传需求合理分配有限地面资源。本论文对卫星数传调度模型及算法进行了研究,主要研究工作如下:卫星数传需求及数传任务是研究卫星数传调度问题的基础,是调度的基本对象。本文提出了卫星数传需求及数传任务概念,建立了卫星数传需求及数传任务模型,规范了卫星数传需求及数传任务的形式化描述。针对可用时间窗口,建立了其基本指标体系,并对不同可用时间窗口间的覆盖关系进行了分析,提出了覆盖度等概念。可用时间窗口概念的提出降低了卫星数传调度问题的求解难度,其基本指标体系及覆盖度等概念为调度过程中可用时间窗口的合理选择提供了依据。本文分析了单数传任务间的冲突,提出了可能冲突等概念,设计了单数传任务集分割算法,给出了可用时间窗口冲突评估指标。可能冲突是针对卫星数传任务特点提出的一个概念,基于可能冲突的单数传任务集分割算法为降低问题规模、降低算法运行时间提供了一种有效方法。可用时间窗口冲突评估指标为调度过程中降低任务之间的相互影响、尽可能调度更多任务提供了支持。本文对卫星数传调度问题中的约束进行了形式化描述,给出了问题模型的一般表示形式及调度目标函数,建立了卫星数传调度问题模型。根据卫星数传调度问题中调度任务集的特点,提出了两阶段调度的问题求解思路,为调度算法的设计提供了基本框架。在对基于综合优先度的卫星数传两阶段调度算法研究过程中,首先提出了综合优先度的概念,并给出了计算公式;然后设计了基于双综合优先度的卫星实拍实传数传调度算法和基于综合优先度的卫星存储转发数传调度算法。这种基于综合优先度的两阶段调度算法实现了两类任务的两阶段调度。基于免疫遗传算法的卫星数传两阶段调度算法是针对卫星数传调度问题设计的另一调度算法。在该算法中,利用免疫遗传算法实现了存储转发数传任务的调度,而对实拍实传数传任务的调度采用了与前一个两阶度调度算法中相同的算法。本文设计了卫星数传调度软件系统,通过设定的调度场景对本文提出的两阶段调度算法的性能及效率进行了验证。

【Abstract】 Satellite data transmission scheduling problem (SDTSP) is assigning limited ground station resources for satellite data transmission requests (SDTR) reasonably. In this paper, scheduling models and its algorithms of STDSP are studied. The main contributions are as follows:As the basic objects of scheduling, SDTR and satellite data transmission task (SDTT) are the basis of researching SDTSP. This paper presents the concepts of SDTR and SDTT firstly, and then constructs their models. These models formalize the description of SDTR and SDTT.Aim to Usable time window (UTW), this paper constructs its basic index system at first, and then analyses the coverage relationship between different UTWs, and brings forward the concept of coverage degree (CD). The use of UTW reduces the complexity of SDTSP. The index system of UTW and CD can be need to support the selection of UTW reasonably during scheduling process.This paper analyses the conflicts between single data transmission tasks, and then presents the concept of possible conflict, designs the algorithm of dividing single data transmission task set, and gives conflict evaluating indexes of UTW. Possible conflict is an important concept presented based on the characteristics of SDTT. A valid method based on dividing algorithm is proposed to reducing the scale of SDTSP and the runtime of scheduling algorithm. The conflict evaluating indexes of UTW provide support for reducing influences between tasks and scheduling as more tasks as possible.Multi-constraint concurrence is the main characteristic of SDTSP. First, the constraints of SDTSP are formalized, and then a general representation form and scheduling object function of SDTSP model are designed. Based on above, this paper builds SDTSP model. According to the characteristic of scheduling task set and for solving SDTSP, a two phases scheduling schema is proposed, which provides a framework for designing scheduling algorithms.During researching on the two phases scheduling algorithm based on synthetical priority of satellite data transmission, this paper presents synthetical priority concept and its computing formulas, and then designs satellite real-time data transmission task scheduling algorithm based on bi-synthetical priority and satellite storage data transmission task scheduling algorithm based on synthetical priority. By employing these two algorithms, this two phases scheduling algorithm achieves two phases scheduling for two kinds of tasks. The two phases scheduling algorithm provides a heuristic method for solving SDTSP.Another two phases scheduling algorithm based on immune genetic algorithm of satellite data transmission is designed for solving SDTSP. In this algorithm, immune genetic algorithm is utilized for scheduling satellite storage data transmission tasks, while the same algorithm as the former two phases scheduling algorithm is used to achieve satellite real-time data transmission tasks scheduling.In the end, a satellite data transmission scheduling software system is designed. Serveral scenarios are designed to show the performance and efficiency of these two phases scheduling algorithms.

节点文献中: