Neuro-physical dynamic load modeling using differentiable parametric optimization
This work addresses the need for accurate load modeling in power systems, but it appears incremental as it builds on existing ZIP models with neural network augmentation.
The paper tackled the problem of creating a reduced equivalent load model for distribution systems in electromechanical transient stability analysis by proposing a neuro-physical model that combines a traditional ZIP load model with a neural network, trained through differentiable programming, and demonstrated its performance on a 350-bus network.
In this work, we investigate a data-driven approach for obtaining a reduced equivalent load model of distribution systems for electromechanical transient stability analysis. The proposed reduced equivalent is a neuro-physical model comprising of a traditional ZIP load model augmented with a neural network. This neuro-physical model is trained through differentiable programming. We discuss the formulation, modeling details, and training of the proposed model set up as a differential parametric program. The performance and accuracy of this neurophysical ZIP load model is presented on a medium-scale 350-bus transmission-distribution network.