This course provides a systems level introduction to modern energy systems and the analytical methods
increasingly used to understand and operate them. The course is organized around three modules:
energy systems, optimization, and artificial intelligence and machine learning. The electric power
system receives particular attention because it connects physical infrastructure, markets, optimization,
data, and AI in a direct and practical way.
Credit Hours: 3
Instructor: Andrew Liu
Email: AndrewLiu@purdue.edu
Syllabus: Coming Soon
Learning Objectives:
• Explain the major components and interactions of modern energy systems, with particular emphasis
on electric power systems and electricity markets.
• Formulate and interpret basic linear and nonlinear optimization models, including power system
operating problems, network constraints, dual variables, and optimality conditions.
• Explain the intuition and mathematical structure behind core supervised, unsupervised, deeplearning, and reinforcement learning methods covered in the course.
•Select and apply appropriate optimization or machine learning tools to representative energy system
problems and interpret the resulting outputs.
• Distinguish what AI methods can contribute to energy systems from what must still be supplied by
physics, data quality, model assumptions, and domain knowledge.
• Discuss how the growth of AI and data centers creates new electricity demand, operational
challenges, and opportunities for grid interaction and flexibility.
Description:
This course provides a systems level introduction to modern energy systems and the analytical methods
increasingly used to understand and operate them. The course is organized around three modules:
energy systems, optimization, and artificial intelligence and machine learning. The electric power
system receives particular attention because it connects physical infrastructure, markets, optimization,
data, and AI in a direct and practical way.
Rather than treating AI as a separate topic layered onto energy, the course develops the connections
among the physical system, optimization models and methodology, and data driven methods. Students
examine both AI for energy -- how optimization and machine learning can support energy system
analysis and decision making -- and energy for AI -- why training large language models and other
foundation models is computationally intensive and requires substantial energy.
Prerequisites:
The course is designed to be largely self-contained. Students should be comfortable with undergraduate calculus, linear algebra, and basic probability and statistics. Prior exposure to optimization, machine learning, or electric power systems is helpful but not required. Basic programming experience, preferably in Python or a comparable language, is expected for computational exercises
Topics:
Module 1: Energy Systems
Module 2: Optimization
Module 3: Artificial Intelligence and Machine Learning
Assignments:
Module1: Brightspace discussion
Module 2: Seven homework assignments
Module 3: Four homework assignemnts, four applied programming assignments
Exams:
Module 1: Quiz
Module 2: One open resource exam
Module 3: One open resource exam
Textbooks:
No required textbook
