Task scheduling of the collaborative aerial–ground system for the search and capture of multiple targets

摘要

Small unmanned rotorcraft (SUR) have been used widely in various civil applications since the opening of the low-altitude airspace. They enable public security maintenance missions via surveilling, searching, and tracking specific areas and targets. SUR functions can be further extended by complementing the ground vehicle and human. In this study, a new collaborative aerial–ground system is introduced to perform a search and capture (SAC) of multiple ground targets. The system comprises several SUR, cars, and humans. The targets primarily refer to criminals or terrorists who tend to escape from the aerial–ground system with holistic and scattered running modes, depending on different situations. A heuristic enumeration optimization (HEO) algorithm is proposed to maximize the probability of detecting the targets in the search space by determining the waypoints for the SUR and cars that can cover additional areas where the targets may occur. Moreover, a dynamic grouping-based task scheduling (DGBTS) algorithm is designed to allocate targets and calculate waypoints for SUR, cars, and humans. In the task allocation problem, the targets are assigned to the SUR by considering the balance of resources, and the quadrant each human is heading for is specified so as to surround the target from different directions. Accordingly, the waypoints are calculated in a distributed approach to highlight the different significance levels of SUR, cars, and humans in the capture phase. Simulation results reveal that the HEO algorithm can detect targets in fewer steps compared to the simulated annealing (SA) algorithm and the asynchronous planning strategy. Moreover, the DGBTS algorithm can deal with different situations in real applications. Additionally, the collaborative aerial–ground system exhibits a better search ability and comprehensive functionality compared to the car–human system and SUR–car systems.