New Paper Alert
✅ Physics-informed backprop for faster AC-OPF learning
✅ Semi-supervised training to slash data prep time
✅ Batch-mean gradient & reduced branches for scalable, real-time-ready training
⚡️ 35× faster inference than solvers
📉 <1% optimality gap
#EnergyAI
🚨 IISE–ESD Energy Data Competition 2026 🚨
Forecast power outages using real-world, high-resolution data.
💰 $1,200 top prize (sponsored by Precision AI LLC)
📊 Data from https://t.co/DWo2mKoHnp
🔗 Details & registration:
https://t.co/b7pX3CVq4T
https://t.co/xuV64keAIm
Join us for an engaging webinar hosted by the IISE Energy Systems Division entitled "Smart Operation of Battery Energy Storage Systems (BESS) in the Evolving Power System" by Dr. Hamid Nazaripouya
Date: Friday, April 25
Time: 2:00 p.m. ET
Register at: https://t.co/ILsWrtypMM
Excited to share our latest work on fault detection in islanded AC microgrids with uncertainties! Our novel observer-centric approach provides robust and responsive fault detection for grid-forming inverters; see https://t.co/dBV57HsLXD
Our paper on load restoration for islanded microgrids was just accepted by @IEEETCNS! Some key aspects:
Model #microgrid physics using DistFlow
Convex relaxations
Tailored constrained policy optimization
https://t.co/yZXyZcq1ih
[New Paper] Inspired by #residuallearning, we propose a physics-guided neural network framework to accelerate probabilistic power flow analysis. Read the full paper here: https://t.co/5Ou9WgynRB
New research improves energy disaggregation using only 1 day of labeled data. Our multi-appliance framework with sample augmentation reduced error by 50%+ for #NILM. #machinelearning#deeplearning
https://t.co/uSEjDu3Uh4
Real-time power system event detection is critical for system stability and resilience. We propose a novel stream learning classifier: the Hoeffding adaptive tree with instance selection for real-time event detection.
https://t.co/o9mfpCbU5y
Check out our recent publication. We propose a unifying deep learning framework for short-term load forecasting. Our novel approach includes time-varying feature weighting, hierarchical temporal attention, and feature-reinforced error correction.
https://t.co/pBvzZyz83Y
Thank you CCST @CCSTorg for having me in this important briefing (https://t.co/RE1FCvu0oG). It's a pleasure & honor to discuss the challenges and opportunities of microgrids. Thank you Chancellor Larive and Dean Wolf @BaskinEng_Dean for the great support! @ucsc@BaskinEng
Communities across California are designing and installing #microgrids to improve their resilience during disasters and to meet their climate change and environmental justice goals.
But what are they, exactly?
🗓️ Join us Wed. 3/30 for a #CCSTbriefing:
https://t.co/y6yUEY3Ubc
Our ECE Dept at UC Santa Cruz is seeking an instructor qualified to teach 175/L Energy Generation and Control and, possibly, other courses within the power electronics research area. Pool applications will be reviewed on March 24, 2022; see https://t.co/NnegwXKzKJ