Enhancing Voltage Compliance in Distribution Network under Cloud and Edge Computing Framework

Jiangxia Zhong,Bin Liu,Xinghuo Yu,Peter Wong,Zeyu Wang,Chongchong Xu,Xiaojun Zhou
DOI: https://doi.org/10.1109/tcc.2022.3149238
IF: 5.697
2022-01-01
IEEE Transactions on Cloud Computing
Abstract:Driven by government incentive policies and heightened environmental awareness by individuals, many regions around the world have seen a rapid rise in distributed energy resource (DER) penetration in electricity distribution networks. While high penetration of DER significantly helps facilitate the decarbonization in power and utitlies, it also brings unexpected operational challenges, among which voltage compliance has been a significant concern. To address this issue, the efficient load profile forecast, the operational framework and related strategies are critical challenges that need to be addressed urgently. Hence, this paper presents a cloud-edge computing-based framework to effectively operate the coupled medium-voltage (MV) and low-voltage (LV) distribution network. The high computational efficiency in cloud computing and low data latency in edge computing are presented and explored to coordinate the day-ahead and intraday operations ranging from different framework layers. Under the framework, a customer-level forecasting algorithm is employed to predict both day-ahead and real-time load profiles. Based on the prediction results, an optimization model based on unbalanced-three phase optimal power flow is proposed and solved by an efficient and accurate linearization-based approach that considers the controllability of on-load tap changers, distributed static var generator and the PV inverters. Simulations based on an extensive mocked MV-LV distribution network show the proposed forecasting method is adopted in real-time operations in terms of high accuracy and demonstrate the efficiency of the proposed optimization method in enhancing the voltage compliance in the network.
computer science, information systems, theory & methods
What problem does this paper attempt to address?