کنفرانس بین المللی کنترل ، ابزار دقیق و اتوماسیون

صفحه اصلی / نهمین کنفرانس بین المللی کنترل ، ابزار دقیق و اتوماسیون

Adaptive Control of Spur Gear Systems via Proximal Policy Optimization and Attention-Based Learning

نویسندگان :

Mohammad Ali Labbaf Khaniki¹ Marzieh Mirzaeibonehkhater² Amirhossein Samii³ Mohammad Manthouri⁴

1- K. N. Toosi University of Technology 2- Indiana University-Purdue University 3- Technical University of Crete Crete, Greece 4- Shahed University

کلمات کلیدی :

deep reinforcement learning،proximal policy optimization،attention-based learning،chaotic systems،adaptive control

چکیده :

This study presents a novel approach for the adaptive control of chaotic spur gear systems using Proximal Policy Optimization (PPO) and attention-based learning. The spur gear system is known for its chaotic behavior, posing challenges for conventional control methods. In this research, the on-policy deep reinforcement learning algorithm, PPO, is employed to effectively control the chaotic dynamics of the spur gear system. PPO is a good choice for controlling chaotic systems because it can learn from the current policy, which is important in situations where the dynamics of the system can change quickly. Furthermore, the neural network architecture within the PPO framework is enhanced by incorporating a self-Attention mechanism. The attention mechanism can adaptively adjust its attention weights based on the current state and policy, allowing the network to dynamically allocate its resources to the most critical aspects of the system. The proposed adaptive control approach demonstrates promising results in effectively managing the chaotic spur gear System, offering potential applications in various real-world scenarios where such systems are encountered.