Scheduling is one of the classic problems in real-time adaptive systems. Due to the complex nature of these applications, the implementation of some sort of run-time intelligence is required, in order to build intelligent systems capable of operating adequately in dynamic environments. The incorporation of artificial intelligence planning techniques in a real-time scenario allows a timely reaction to external and internal events. In this work, a layered architecture integrating real-time scheduling at the bottom level and artificial intelligence planning techniques at the top level has been designed, to implement a multi-level scheduler with the capability to perform effectively in this kind of situation. This multi-level scheduler has been implemented and evaluated in a simulated information access system destined to broadcast information to mobile users in a time-constrained communication environment, modeling mobile users’ realistic information access patterns. Results show that the incorporation of artificial intelligence planning improves the overall performance, adaptiveness, and responsiveness with respect to the non-AI-based scheduler version of the system.

调度是实时自适应系统中的经典问题之一。由于这些应用程序的复杂性，因此需要某种形式的运行时智能，以构建能够在动态环境中充分运行的智能系统。将人工智能计划技术结合到实时场景中，可以及时对外部和内部事件做出反应。在这项工作中，已经设计了一个集成了底层底层实时调度和顶层人工智能规划技术的分层体系结构，以实现具有在这种情况下有效执行的能力的多层调度程序。此多级调度程序已在旨在在时间受限的通信环境中向移动用户广播信息的模拟信息访问系统中实现和评估，从而对移动用户的现实信息访问模式进行了建模。结果表明，相对于基于非AI的计划程序版本，人工智能计划的合并可提高整体性能，适应性和响应能力。