论文部分内容阅读
对于采用分层的主从Agent体系结构的航天器自主运行系统而言,各个子Agent可以单独设计,这样就大大减小了设计难度并提高了程序运行效率。对于每个子Agent而言,可以专注于底层模型和算法的设计。空间观测任务调度子Agent的任务是对空间观测任务进行调度,以使在有限的资源下,可观测任务最多,观测价值最大。针对这一子Agent,建立了优化模型来完成最优调度问题的建模。优化模型的目标函数为观测任务的优先级之和最大,优先级与观测任务的价值有关。约束条件包括观测机会约束、航天器资源约束等。采用基于遗传算法的启发式算法进行求解,建立了具有可扩展性的任务调度仿真演示系统,直观的演示了调度后的任务运行情况。
For the autonomous system of spacecraft with hierarchical master-slave architecture, each sub-agent can be designed separately, which greatly reduces the design difficulty and improves the program operating efficiency. For each sub-agent, you can focus on the design of the underlying models and algorithms. The task of the space observation task scheduling sub-agent is to schedule the space observation task, so that under the limited resources, the most observable task and the maximum observation value are. For this sub-agent, an optimization model is established to complete the modeling of the optimal scheduling problem. The objective function of optimization model is that the sum of priority of observation task is the highest, and the priority is related to the value of observation task. Constraints include observation opportunity constraints, spacecraft resource constraints. The heuristic algorithm based on genetic algorithm was used to solve the problem. A scalable task scheduling simulation demonstration system was established, demonstrating the operation of the scheduled task visually.