Nonlinear model reduction of dynamical power grid models using quadratization and balanced truncation

TL;DR

This paper introduces a nonlinear model reduction technique for power grid models by reformulating them as quadratic systems and applying balanced truncation, enabling faster real-time security assessments.

Contribution

It presents a novel quadratic reformulation and balanced truncation method for reducing large-scale nonlinear power grid models, improving computational efficiency.

Findings

Successful reduction of IEEE 57 and 118 bus system models

Reduced models maintain essential dynamic characteristics

Demonstrated potential for real-time security analysis

Abstract

In this work, we present a nonlinear model reduction approach for reducing two commonly used nonlinear dynamical models of power grids: the effective network (EN) model and the synchronous motor (SM) model. Such models are essential in real-time security assessments of power grids. However, as power grids are often large-scale, it is necessary to reduce the models in order to utilize them in real-time. We reformulate the nonlinear power grid models as quadratic systems and reduce them using balanced truncation based on approximations of the reachability and observability Gramians. Finally, we present examples involving numerical simulation of reduced EN and SM models of the IEEE 57 bus and IEEE 118 bus systems.