Edge Computing Resource Optimal Dispatch in Distribution Network Considering the Power Tasks Uncertainty

With the increasing penetration rate of new energy in the power grid, the tasks scenarios faced by the power system are becoming increasingly complex. This complexity leads to the spatiotemporal uncertainty about resource requirements of terminal task for edge computing. Traditional resource scheduling schemes are configured with constant values or over-provisioned in each power business scenario, resulting in low resource utilization. A data-driven model to optimize scheduling resources, which can be for edge computing terminals in distribution network, is proposed to meet the uncertain resource requirements. Firstly, a power tasks resource demand model considering source-load uncertainty is established, and the boundary conditions of adjustable resources are established by utilizing the adjustability of power tasks with different delay and priority requirements. Secondly, considering the multiple costs consumed in the process of power tasks scheduling, a two-stage optimization model is established with the comprehensive computing resource balance and operation cost as the objective function. Finally, the simulation results show that the proposed model has a positive effect on improving the balance and economy of distribution network edge computing resources compared with the traditional optimization model.