.png){kind=icon}
.png){kind=icon}
Return
Distributed Task Allocation Algorithm for Heterogeneous UAVs Based on Reinforcement Learning
.png){kind=icon}
OA
AI
DOI：10.3390/drones10030220.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Highlights What are the main findings? The proposed decentralized reinforcement learning algorithm, integrated with the PPO and a LSTM-attention fusion network, eliminates reliance on a global scheduler and addresses sparse reward instability, achieving 100% task success rate across diverse scenarios. The Cascaded Commitment Timeout (CCT) mechanism effectively prevents training deadlocks, while the fusion network captures temporal and collaborative dependencies, enabling zero-shot generalization and superior robustness against UAV failures and dynamic tasks. What are the implications of the main findings? Theoretical Implication: This study enriches the family of distributed multi-agent task allocation methods, providing a viable framework for addressing deadlock and temporal dependency challenges in reinforcement learning-based scheduling. Practical Implication: The algorithm's high robustness and zero-shot adaptability make it suitable for real-world applications like low-altitude logistics and emergency rescue, supporting flexible scaling of UAV fleets and task volumes.Highlights What are the main findings? The proposed decentralized reinforcement learning algorithm, integrated with the PPO and a LSTM-attention fusion network, eliminates reliance on a global scheduler and addresses sparse reward instability, achieving 100% task success rate across diverse scenarios. The Cascaded Commitment Timeout (CCT) mechanism effectively prevents training deadlocks, while the fusion network captures temporal and collaborative dependencies, enabling zero-shot generalization and superior robustness against UAV failures and dynamic tasks. What are the implications of the main findings? Theoretical Implication: This study enriches the family of distributed multi-agent task allocation methods, providing a viable framework for addressing deadlock and temporal dependency challenges in reinforcement learning-based scheduling. Practical Implication: The algorithm's high robustness and zero-shot adaptability make it suitable for real-world applications like low-altitude logistics and emergency rescue, supporting flexible scaling of UAV fleets and task volumes.Abstract To address the challenges faced by heterogeneous Unmanned Aerial Vehicle (UAV) systems in complex task allocation, including over-reliance on centralized scheduling, training deadlock, inadequate capture of temporal collaboration, and unstable training under sparse reward conditions, this paper proposes a distributed task allocation algorithm based on reinforcement learning. The algorithm adopts a decentralized decision-making architecture, which enables the autonomous formation of UAV collaborative groups without the need for a global scheduling center. A cascaded submission timeout mechanism is introduced to prevent training deadlock; the combination of Long Short-Term Memory (LSTM) and attention mechanism is employed to accurately model temporal correlations and collaborative dependencies; and the Proximal Policy Optimization (PPO) algorithm is leveraged to optimize the training stability under sparse reward conditions. Experimental results demonstrate that the proposed algorithm achieves a 100% task success rate in scenarios of different scales, and its key metrics, including makespan, time cost and waiting time, are significantly superior to those of mainstream baseline methods such as the Genetic Algorithm (GA) and the Hungarian Algorithm (HA). Moreover, the algorithm still maintains excellent robustness under the conditions of UAV failures, parameter variations, and dynamic task perturbations. This method supports zero-shot generalization for any number of UAVs and tasks and provides an efficient and reliable solution for the real-time collaborative scheduling of heterogeneous UAV systems.
Keywords:
UAV
task allocation
reinforcement learning
encode
Proximal Policy Optimization