变电站无人机机巡边缘计算框架设计及资源调度方法

Framework Design and Resource Scheduling Method for Edge Computing in Substation UAV Inspection

变电站无人机巡检时,对计算和存储资源进行科学调度有助于快速分析设备缺陷。针对无人机巡检图像分析需要的计算资源及存储资源,基于工业物联网构建了包括数据分析服务器、无人机、物联网终端设备在内的2层边缘计算框架模型;根据变电站机巡的特点,建立了面向任务以及待巡检设备的数据模型及巡检流程;应用Stackelberg多层博弈算法,以计算量为资源调度的定价依据,将物联网终端自身的数据资源和计算资源作为分配资源的约束条件,并从理论上证明效用函数满足凸函数条件,也存在纳什均衡点,并通过仿真验证了无人机与数据分析服务器最优价格曲线交点即为资源调度的纳什均衡点。实际应用结果表明,变电站数据网络传输以及计算量大幅度降低,全站巡检时间由15 h下降为12 h。该方法可以为今后推进高效、安全的变电站无人化运维提供依据。

Scientific scheduling of computation and storage resources can be helpful to analyzing equipment defects quickly during substation UAV inspection.The computation resources and storage resources are required in the process of analyzing UAV inspection images,thus a two-layer edge computing framework model including data analysis servers,UAVs and terminal equipment of the internet of things is built based on the industrial internet of things(IoT).According to the characteristics of substation inspection,a data model of task-oriented and equipment to be inspected is established,the resource state is analyzed,and the control flow of inspection is proposed.By using Stackelberg multilayer game algorithm and computing volume as the pricing basis of resource scheduling,the data resources and computation resources of the terminal of IoT are taken as the constraints for resource allocation.Theoretically,it is proved that the utility function satisfies the convex function condition and the Nash equilibrium point exists.In the simulation analysis,the change of task allocation and the relationship between utility function,revenue function and calculation allocation during iteration are analyzed.The results reveal that the intersection point of optimal price curves of UAV and data analysis server is the Nash equilibrium point of resource scheduling.The practical application results show that the transmission and calculation amount of substation data network are greatly reduced,and the inspection time of the whole station reduces from 15 hours to 12 hours.This method can provide the basis for the efficient and safe operation and maintenance of unmanned substation.