Offline Task Assignment and Motion Planning Algorithm Considering Agent's Dynamics

Abstract

The aim of this study is to develop an algorithm for task assignment and movement planning to a destination by considering the acceleration and deceleration patterns of automated guided vehicles (AGVs). In this study, we propose an algorithm for task assignment and motion planning for AGV kinematics with multiple search trees based on the conflict-based search with optimal task assignment (CBS-TA) of Hönig et al. By devising the options for AGV actions and settings for standby actions, the algorithm can perform optimal motion planning for each AGV, considering its dynamic characteristics.