Distributed Assistive Control of Power Buffers in DC Microgrids

Abstract

Low generational-inertia and lack of damping elements cause stability concerns in DC microgrids. Power buffers have been introduced to damp volatile load demands and improve microgrid’s stability. Power buffer is a power electronics converter with large storage components (e.g., capacitors), that can decouple dynamics of power distribution network and electronic loads by adjusting its input impedance and stored energy. Typically, power buffers are controlled individually to serve local loads. Alternatively, collective operation of power buffers is considered here to extend their damping effect to neighboring loads. Additionally, the group operation allows buffer designs with smaller storage components. A supervisory control to manage energy-impedance profiles of power buffers across a grid would require a complex communication network. Alternatively, distributed control lays a reliable ground to link power buffers with a minimal communication. This work offers a fully distributed feedback control algorithm to collectively manage the power buffers, using a sparse communication network. The controller enables the buffers to collectively respond to any load transient. Hardware-in-the-loop simulation of a DC microgrid is used to show the controller’s efficacy; it successfully groups power buffers in the neighborhood of the affected load and manages their energy reservoirs to shape the power supplied by the grid.

Existing System

Power Factor-Correction Scheme, Optimization.

Proposed System

This paper proposes a distributed feedback control paradigm that groups the buffers through a sparse communication network. The grouped buffers collectively respond to load changes all across the microgrid; as opposed to individual actions where each buffer would respond only to its local load change. Potentially, this collective behavior enables buffer designs with smaller energy capacity to be as effective as larger designs. Salient features of this work are as follows: Each load is augmented with a power buffer; the buffer is placed between the load and the point of connection to the distribution network. Incorporating the power buffer in the load structure adds a degree of control freedom to enhance the system’s transient response. Each buffer exchanges data with only a few other buffers neighboring on a sparse communication graph. These distributed control methods are well-established to be superior over centralized methods. While fully-decentralized droop mechanisms are prevalent in both AC and DC microgrids, the advantages of distributed solutions, with limited message passing among participating agents on a sparse communication network. In contrast to majority of distributed control approach in the literature that focus on the source-side management, the proposed method focuses on the collective energy management of power buffers at the load side. The proposed controller processes local and neighbors’ voltage and power measurements to collectively decide the desired energy profile of the buffer during any transient. The desired energy profile is then translated into an equivalent voltage profile that is tracked by the buffer. While power information directly updates neighbors’ power demand, voltage data is chosen as the secondary variable to help sensing any load change. This model plays a pivotal role in designing, evaluation and tuning the controller performance.

Architecture

Proposed Control Methodology