Applied AI for Power and Energy Systems
ECE480A9 · Colorado State University
Fall 2026 | Fort Collins, CO
Why I took it
At BOR I sit next to hydropower plants and a digital exciter — plant data, frequency response, and a controller that has to be right when it ships. This course pairs each AI method with a grid problem (forecasting, faults, state estimation, frequency control) instead of teaching models in the abstract. That is the version I wanted.
What I'm learning
ML on power-system data
Core ML on engineering models, then unsupervised methods for fault and anomaly detection, and time-series models for load and renewable generation forecasting.
Structure-aware models
CNNs for contingency and security assessment, graph networks for state estimation, physics-informed networks for optimal power flow, and neural ODEs for system dynamics.
Learning for grid control
Reinforcement learning for frequency response, multi-agent RL for distributed energy resource coordination, and Bayesian methods for risk-aware operations.