Decentralized Model Predictive Control of Hybrid Distribution Transformers for Voltage Regulation in Active Distribution Networks

Abstract

Hybrid distribution transformer (HDT) can provide dynamic ac voltage regulation at the distribution level. In order to take advantage of this capability, this article presents a decentralized model predictive control scheme, which includes HDTs in the active distribution network (ADN) voltage control. This scheme employs a fully decentralized architecture, in which each HDT is controlled by a local robust MPC (RMPC). Each local RMPC, based solely on local measurements, computes the optimal reactive power reference for its associated HDT. The optimization aims at correcting abnormal voltages and reducing reactive power injection/absorption. By doing so, this scheme achieves both objectives of ADN voltage regulation and control effort minimization. The salient feature of this scheme is that, by regarding the interactions among HDTs as disturbances and compensating them using tube-based robust method, it achieves a suboptimal control of ADN bus voltages without communication among HDTs. Simulation results verify the effectiveness of the proposed scheme.