Extensions of Receding Horizon Task Assignment for Area Coverage in Dynamic Environments

Abstract

This study presents a task reassignment strategy that can flexibly handle changes in dynamic environments. Most existing studies have only considered static environments where the mission is predetermined and unexpected events do not occur. However, in practice, the environment in which unmanned aerial vehicles perform their mission is complex and includes various pop-up events. Therefore, task reassignment is essential through re-optimization to adjust the previous plan, which is no longer optimal whenever dynamic events occur. The importance of this study is that the modified receding horizon task assignment algorithm is applied for accurate and efficient task reassignment. In addition, this study aims to provide a complete system architecture from operator input to motor commands for autonomous area coverage missions. Reliable performance validation is also performed through various simulations and a challenging outdoor flight experiment with real hardware implementation.

A multi‐UAV system for coverage path planning applications with in‐flight re‐planning capabilities

“…In this case, it can be observed that the planning times do not depend on the number of UAVs (as shown in Figure 12); this is because we perform the simulation tests with random area regions, which affects the CPU time. In the dynamic re‐planning approach presented by (Hong et al, 2022) the authors use a a 2.6 GHz Intel i7 CPU and 16 GB RAM to perform their experiments, obtaining from 260 to 710 ms in each mission replanning while our method obtain a maximum of 14 ms in a CPU with similar features. In this context, our system maintains mission continuity by handling shorter replanning times.…”

“…For instance, works like Chen et al (2022), Meng et al (2009) study the task replanning problem in dynamic environments, recalculating the optimal path in certain events, these investigations determine the shortest route for exploring multiple waypoints and replanning in case of finding new information, but they do not address the area coverage problem. Moreover, the research presented in Hong et al (2022) address the CPP problem with in‐flight replanning using a heuristic algorithm called RHTA. In this architecture, two types of replanning events are handled.…”

A multi-UAV system for coverage path planning applications with in-flight re-planning capabilities

“…Tables 7 and 8 show the low computational time of the proposed algorithms during planning and re-planning tasks. Compared with the dynamic re-planning approach presented by (Hong et al, 2022) At a second time, UAV2 fails, and UAV1 recovers, so the system replans the mission for UAV1 and UAV3, taking over the paths for UAV2. Figure 13 also shows the altitude transitions to ensure the safety given by the Powell-BINPAT algorithm.…”

“…For instance, works like (Chen et al, 2022;Meng et al, 2009) study the task replanning problem in dynamic environments, recalculating the optimal path in certain events. Additionally, the research presented in (Hong et al, 2022) uses a heuristic algorithm called RHTA for CPP with dynamic replanning in a centralized architecture. Alternatively, to above mentioned methods for single UAV area coverage, UAV software architecture for multiple aircraft have captured the attention of the UAV swarm intelligence scientists (Zhou et al, 2020).…”