Optimization-Based Control of Distributed Battery Storage in Distribution Networks

TL;DR

This paper introduces a combined global-local control method for distributed energy resources in power distribution networks, improving voltage regulation and reducing power losses through an optimization-based approach.

Contribution

It presents a novel convex multi-period optimization framework that updates local controller gains, ensuring stability and respecting DER constraints.

Findings

Voltage regulation improved significantly.

Power losses reduced by 11%.

Peak substation power decreased by 26%.

Abstract

We propose a combined global-local control approach to regulate voltage and minimize power losses in distribution networks with high integration of distributed energy resources (DERs). Local controllers embed the fast acting proportional volt-var-watt control law and have their gain (slope) coefficients updated regularly by a global optimization problem at a slower time-scale. Design of optimal coefficients preserve overall system stability and encapsulate inverter and energy limits of controllable DERs. The proposed approach is formulated based on a linear network model (LinDistFlow) and suitable approximations to produce a convex multi-period optimization formulation. Numerical simulations with real-world customer data and two different distribution feeders revealed that our approach provides substantial voltage regulation, while reducing losses by 11 per cent and peak substation power by 26 per cent compared to other state-of-the-art algorithms.