Improved CBBA Algorithm for Task Allocation in Multi-UAV Inspections of Power Systems

The technology of multi-UAV coordination and formation flight has a broad prospect in the field of power system inspection. Aiming at the collaborative operation of multiple UAVs in foreign object removal tasks during power inspection, this paper proposes an improved CBBA algorithm to address the task allocation problem of heterogeneous UAVs in dynamic and time-sensitive environments. The algorithm integrates the distributed structure of market mechanisms and the convergence idea of consensus strategies, which is particularly suitable for dealing with time-varying and dynamic task allocation scenarios. This paper proposes an optimized CBBA algorithm that improves the communication structure and reward function, introduces and optimizes the CBBA-LR algorithm to handle task conflicts, and uses a grouped multithreading pre-treatment mechanism to reduce computational load. These improvements enable the algorithm to ensure the reasonableness of task allocation while guaranteeing the efficiency of search and rescue, and can quickly provide effective task allocation schemes for large-scale UAV swarms. Simulation results verify the advantages of the improved CBBA algorithm in improving task allocation efficiency and reducing computational pressure.