This work is the first to adopt Kolmogorov-Arnold Networks (KAN), a recent breakthrough in artificial intelligence, for smart grid optimizations. To fully leverage KAN's interpretability, a general framework is proposed considering complex uncertainties. The stochastic optimal power flow problem in hybrid AC/DC systems is chosen as a particularly tough case study for demonstrating the effectiveness of this framework.
@article{arxiv.2408.04063,
title = {From Black Box to Clarity: AI-Powered Smart Grid Optimization with Kolmogorov-Arnold Networks},
author = {Xiaoting Wang and Yuzhuo Li and Yunwei Li and Gregory Kish},
journal= {arXiv preprint arXiv:2408.04063},
year = {2024}
}
Comments
Accepted in Late Breaking Research Publications in 2024 IEEE Energy Conversion Congress and Exposition (ECCE)