On February 28, 2025, at 17:15, during our seventh seminar of this year, Milena Ispiryan, a research intern at the Center for Scientific Innovation and Education, presented the summarized results of her research on the topic "Inverted Pendulum: A RL Approach"

The seminar covers:

• The fundamental principles of RL and its application in control systems,
• Reward-based learning,
• The application of RL in real-world control systems.

Key Highlights:

✅Designed and modeled an RL-based control system for the Inverted Pendulum,

✅Trained a deep learning model using the Deep Deterministic Policy Gradient (DDPG) method,

✅Evaluated the system's stability and efficiency,

✅Compared RL results with classical Linear Quadratic Regulator (LQR) control methods.

This research demonstrates how AI-based controllers can autonomously learn to stabilize unstable systems, paving the way for new advancements in automation and robotics.