Power Loss Minimization of Parallel-Connected Distributed Energy Resources in DC Microgrids Using a Distributed Gradient Algorithm-Based Hierarchical Control

Abstract

This paper presents a distributed hierarchical control to minimize the distribution power loss, including both line loss and converter loss, of distributed energy resources (DERs) in DC microgrids. The proposed control comprises a distributed gradient algorithm (DGA) at the top layer, a consensus control at the secondary layer, and a local droop and dual-loop control at the bottom layer. Since the distribution power loss of DERs is modelled a concave function of the output currents, the DGA with addition and subtraction operators can be directly used to derive the optimal current allocation coefficients based on peer-to-peer data exchange between the neighboring DERs. The derived optimal current allocation coefficients are adopted by the distributed secondary control to generate adaptive voltage terms for the local control. The local control consists of a droop control and a classic dual-loop control for grid-connected converters. Both simulation and experimental results have validated that the proposed control can reduce more distribution power loss of DERs than the conventional control by only considering line loss by various cases, including normal operation, power clip, plug-in-and-out of DERs, and communication failure. Experimental results also verify that the execution speed of the proposed control is faster than that of the conventional Lagrange multiplier-based centralized control.