Affiliations: [a] School of Management, Huazhong University of Science and Technology, Wuhan, China
| [b] School of Automation, Huazhong University of Science and Technology, Wuhan, China
| [c] MOE Key Laboratory of Image Processing and Intelligent Control, Huazhong University of Science and Technology, Wuhan, China
Abstract: Emergency decision-making confronts a problem where emergency tasks coordinating to achieve emergency objectives are subject to urgent time and limited resources. Actions with uncontrollable durations make the problem more complicated. This paper proposes a novel resource-constrained Hierarchical Task Network planning approach under uncontrollable durations for emergency decision-making. The objective is to generate a dynamically controllable plan with constrained resources and execute the plan via scheduling actions dynamically. Two of the most general categories of resources in emergency decision-making, consumable resources and reusable resources, are considered. First, timed initial literals are extended to present timed resources and the constraints of timed literals are transformed to impose restrictions on actions in planning. Second, a mechanism is explored to hierarchically inherit the resource constraints of compound tasks by subtasks during Hierarchical Task Network planning. Third, two categories of resource constraints are checked, respectively. Potential resource conflicts are resolved via adding precedence constraints. Finally, experimental studies are conducted to evaluate the planning approach. The results demonstrate the effectiveness and efficiency of the planning approach for emergency decision-making.
