Distributed economic nonlinear MPC for DC Micro-Grids with inherent bounded dynamics and coupled constraints

This paper proposes a hierarchical control scheme for DC Micro-Grids (MGs) composed of distributed energy resource (DER) units, active loads, and coupled constraints, that incorporates different control levels to achieve economic operation and coupled constraint satisfaction at all times. All loads are considered as constant power loads (CPLs), which are known to lead to instability due to the negative impedance characteristics. To this end, firstly an improved version of the recently developed sl-PI controller is introduced in the primary level to provide boundedness and local stability of the dynamics. The supervisory controller solves an economic optimisation problem to compute voltage setpoints and at a lower control layer the system is driven to those setpoints, by utilising a distributed nonlinear MPC setup with non-iterative, neighbour-to-neighbour communication framework. A rigorous analysis on the system robustness to uncertainties in the exchanged information and bounded load variations is provided, and recursive feasibility and stability of each subsystem is analytically proven for the first time, according to the authors' knowledge. Finally, we demonstrate the proposed controller with the aid of simulated scenarios for a DC MG with multiple DER units and CPLs. (c) 2022 Elsevier B.V. All rights reserved.

Automated summary

This paper proposes a hierarchical control scheme for achieving economic operation and coupled constraint satisfaction in DC Micro-Grids (MGs) composed of distributed energy resource (DER) units, active loads, and coupled constraints. The control scheme incorporates different control levels, including an improved version of the sl-PI controller, a supervisory controller for economic optimization, and a distributed nonlinear MPC setup. Rigorous analysis is provided to demonstrate the robustness of the system to uncertainties and load variations. Simulated scenarios are presented to demonstrate the effectiveness of the proposed controller.