Optimization of Task Scheduling Strategy for UAV-enabled Cloud-edge Collaboration System using Reinforcement Learning Method

Transmission line inspections have increasingly adopted UAVs equipped with multiple sensors to collect data during inspection processes. However, the limited computational capabilities of UAVs make processing multiple data types simultaneously challenging, often leading to high processing latency. To address this issue, this paper proposes an optimized UAV task scheduling strategy based on cloud-edge collaboration. The proposed strategy utilizes reinforcement learning to enable dynamic offloading of computational tasks to edge servers and cloud servers, thereby reducing task execution time and enhancing overall inspection efficiency. Experimental results demonstrate the effectiveness of the proposed approach in minimizing processing latency and improving the reliability of transmission line inspections.