Leveraging Nonsmooth Barrier Functions Within Behavior Trees for Managing Multirobot Task Conflicts in Persistent Coverage

摘要

Conflicts may arise in multirobot systems when several tasks are executed concurrently. We propose a generalized nonsmooth barrier function-behavior tree (NBF-BT) method. It aims to address conflicts arising from subtask constraints during persistent coverage tasks. This method integrates nonsmooth barrier function (NBF) concepts into the behavior tree (BT) method, using the reactivity of the BT to dynamically adjust the set of multiple barrier functions. It applies a set of barriers to distributed multirobot systems, thereby enhancing overall task performance. Specifically, the barrier function modifies the leaf nodes of the BT, serving as a task manager for each robot. During mission performance, it can dynamically add or remove constraint functions from the constraint set based on current environmental conditions and the states of other robots. When multiple tasks assigned to a robot conflict with each other, this adaptive mechanism sacrifices some local performance to boost the overall performance of the multirobot system, effectively resolving task conflicts within the robotic team. Moreover, this method can restore the system to its initial state following conflict resolution, exhibiting robust performance. To verify the effectiveness of our method, we apply the proposed approach to the complex task of persistent coverage and surveillance carried out by multirobot, addressing various constraints in task conflict. We conducted two sets of simulation experiments a to evaluate the efficacy of the proposed method in managing multirobot conflict tasks, evaluating the coverage capability of the multirobot system within the environment. a https://youtu.be/h7K3wIKQIKc .