Multi-UAV Multi-Task Allocation Based on Cognitive State-Based Hybrid Contract Net Protocol

The future of Multi-UAV systems will evolve towards being more clustered, autonomous, and intelligent. By integrating information and complementing capabilities, these systems can overcome individual limitations, forming diverse task capabilities to meet the needs of future warfare. Multi-drone task allocation, as a critical design aspect of multi-drone collaborative applications, is essentially a constrained combinatorial optimization problem aimed at obtaining feasible task allocation solutions. To address the dynamic task scheduling problem in multi-drone systems, this paper establishes a multi-constrained mathematical model that incorporates task requirements and platform capabilities. Three objectives are considered: maximizing the total profit of scheduled tasks, minimizing time consumption, and balancing the number of tasks assigned to each drone. A task scheduling method based on a hybrid contract net protocol with cognitive state considerations is proposed, where task types include buy-sell contracts, exchange contracts, and replacement contracts. The cognitive ability state of the drones is dynamically updated. Finally, extensive simulation experiments are conducted under dynamic scenarios, including urgent tasks and unforeseen obstacles, to validate the superiority of the proposed method.