Robustly Coordinated Distributed Voltage Control Through Residential Demand Response Under Multiple Uncertainties

Abstract

This article proposes a multilevel coordinated voltage control (CVC) strategy for facilitating distributed residential demand response for network voltage optimization. The proposed CVC scheme coordinates multiple home energy management systems (HEMS) connected on a low-voltage radial distribution network through a sensitivity-based approach to minimize voltage deviations in multiple timescales owing to high photovoltaic (PV) penetrations and load variability. In order to mitigate the effects of uncertain PV output and load during actual system operation, a day-ahead two-stage robust optimization (TSRO) model has been developed, which optimizes network operation in the longer timescale, for worst-case realization of uncertainties in shorter timescales. This is sequentially followed by fully distributed hour-ahead CVC which is decomposed into network optimization and energy management subproblems through alternating direction method of multipliers. In order to counteract the effects of real-time uncertainties, local adjustments are performed by redispatching HEMS at individual households to locally adapt to unforeseen variations, thereby minimizing real-time voltage deviations. Extensive simulations are performed on the modified Dutch LV network. Simulation results demonstrate effectiveness of the proposed multilevel CVC approach to robustly handle uncertainties while ensuring reliable operation and maximum user-comfort.