Summary of Function Approximation For Reinforcement Learning Controller For Energy From Spread Waves, by Soumyendu Sarkar et al.
Function Approximation for Reinforcement Learning Controller for Energy from Spread Waves
by Soumyendu Sarkar, Vineet Gundecha, Sahand Ghorbanpour, Alexander Shmakov, Ashwin Ramesh Babu, Avisek Naug, Alexandre Pichard, Mathieu Cocho
First submitted to arxiv on: 17 Apr 2024
Categories
- Main: Artificial Intelligence (cs.AI)
- Secondary: Machine Learning (cs.LG); Systems and Control (eess.SY)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary A novel Multi-Agent Reinforcement Learning (MARL) controller, trained with the Proximal Policy Optimization (PPO) algorithm, is developed to efficiently capture energy from spread waves in industrial multi-generator Wave Energy Converters (WECs). The MARL controller addresses multiple objectives simultaneously, including energy capture efficiency, reduction of structural stress, and proactive protection against high waves. The paper explores different function approximations for policy and critic networks to model the sequential nature of system dynamics, finding that they are crucial for improved performance. Experimental results demonstrate that a transformer model with gated residual connections achieves optimal performance, boosting energy efficiency by 22.1% compared to traditional spring damper (SD) controllers. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Wave Energy Converters (WECs) need special controllers to handle many waves coming from different directions. This paper shows how to use Multi-Agent Reinforcement Learning (MARL) to make a better controller. The MARL controller is trained using an algorithm called Proximal Policy Optimization (PPO). It does three things well: captures energy, reduces stress on the device, and protects it from big waves. The researchers tested different ways of designing the controller’s policy and critic networks and found that they are important for good performance. They also tried out a special kind of neural network called a transformer model and found that it worked really well. |
Keywords
» Artificial intelligence » Boosting » Neural network » Optimization » Reinforcement learning » Transformer
