Autonomous power management and load sharing in isolated micro-grids by consensus-based droop control of power converters

Autonomous power management and load sharing is one of the important aspect in an isolated micro-grid. Conventionally, this task was achieved by the local real power-frequency and reactive power-voltage droop control of individual distributed energy resources (DERs) interface power converters (DICs). Even though this method is simple, some flaws about reactive power sharing and voltage restoration mechanism have been reported recently. In order to enhance this conventional droop control method, various approaches have been proposed. One possible solution is to provide communications networks among power converters. Under this framework, the control structure of a micro-grid can be represented as a multi-agent system. In this paper, the consensus-based droop control method will be developed for autonomous power management and load sharing. Two variables which reflect the average consensus about frequency deviations and voltage magnitude deviations will be appended in the power control loop of each power converter. By exploring transient energy functions of the bulk micro-grid, it will be shown that both real power and reactive power can be properly shared even for multiple loads under large disturbances if all line impedances have the same R/X ratio. The advantage of the proposed consensus-based control will be discussed in details. Under this framework, the closed-loop system can be proven to be quasi-gradient and the stability can be indeed improved by providing additional damping from consensus-based control. Dynamical simulations of a 6-DIC/14-bus micro-grid are performed under the real-time simulation environments OPAL-RT with detailed dynamical converter models. Satisfactory results will be provided to validate the effectiveness of the proposed consensus-based control method.