Fast Temporal Logic Mission Planning of Multiple Robots: A Planning Decision Tree Approach

Abstract

This work develops a fast mission planning framework named planning decision tree (PDT), that can handle large-scale multi-robot systems with temporal logic specifications in real time. Specifically, PDT builds a tree incrementally to represent the task progress. The system states are modeled by both completion positions and times, which avoids sophisticated product automaton and significantly reduces the search space. By growing the tree from the root node to leaf nodes, PDT can be searched for mission plannings that satisfy the linear temporal logic (LTL) task. Rigorous analysis shows that the PDT based planning is feasible (i.e., the generated plan is applicable and satisfactory with respect to the LTL task) and complete (i.e, a feasible solution, if exits, is guaranteed to be found). We further show that PDT based planning is efficient, i.e., the solution time of finding a satisfactory plan is only linearly proportional to the robot numbers. Extensive simulation and experiment results demonstrate its efficiency and effectiveness.