Hierarchical mission replanning for multiple UAV formations performing tasks in dynamic situation

Abstract

UAV should be capable of responding to the dynamic changes, timely and accurately. The mission replanning problem for multiple UAV formations is studied in this paper, and a hierarchical method is proposed to have a comprehensive consideration of the dynamic changes caused by the task and environment. First, three types of task change are analyzed, and the task reassignment approach is designed to determine the opportunity of conducting path replanning, which can make the best of the predetermined path and save more computational time. Moreover, a hierarchical strategy is presented to determine the most suitable UAV formation to perform a new task. Then a two-step path replanning approach combining the efficient rapid-exploration random tree (ERRT) algorithm and the modified consensus (MC) algorithm is developed. The waypoints are determined by the ERRT algorithm, and then the MC algorithm is applied to make the UAV formation head for the waypoints one after another. The MC algorithm also can be used in the UAV formation’s dismission and a single UAV’s landing. Simulation results demonstrate that the proposed hierarchical mission replanning method performs well under complicated dynamic changes. The task list of UAV formation can be updated timely and reasonably, and feasible paths can be generated rapidly by the ERRT algorithm considering different pairs of start point and destination.