Linear Aggregate Model for Realizable Dispatch of Homogeneous Energy Storage

TL;DR

This paper introduces a linear aggregate model for homogeneous energy storage systems that enables convex optimization of battery dispatch while guaranteeing the realizability of the power schedule, addressing the non-convexity challenge.

Contribution

It develops a novel linear composition of energy storage elements that allows independent charging/discharging and convex optimization of the aggregate battery.

Findings

Enables convex optimization of battery dispatch.

Guarantees admissibility of power schedules.

Supports disaggregation to individual storage elements.

Abstract

To optimize the dispatch of batteries, a model is required that can predict the state of energy (SOE) trajectory for a chosen open-loop power schedule to ensure admissibility (i.e., that schedule can be realized). However, battery dispatch optimization is inherently challenging when batteries cannot simultaneously charge and discharge, which begets a non-convex complementarity constraint. In this paper, we develop a novel composition of energy storage elements that can charge or discharge independently and provide a sufficient linear energy storage model of the composite battery. This permits convex optimization of the composite battery dispatch while ensuring the admissibility of the resulting (aggregated) power schedule and its disaggregation to the individual elements.