Recurrent Soft Actor Critic Reinforcement Learning for Demand Response Problems

Keyphrases

• Actor-critic Reinforcement Learning 100%
• Demand Response 100%
• Soft Actor-critic 100%
• Control Solution 60%
• Model Predictive Control 40%
• Policy Function 40%
• Recurrency 40%
• Reinforcement Learning Algorithm 20%
• Continuous Action 20%
• Performance Improvement 20%
• Rule-based 20%
• Problem-based 20%
• Prediction Accuracy 20%
• Actor-critic Methods 20%
• Markov Decision Process 20%
• Rule-based Model 20%
• Discrete Actions 20%
• Rule-based Control 20%
• Heat Pump 20%
• Partially Observable 20%
• Non-recurrent 20%
• Action Capabilities 20%
• Complete Observability 20%
• Alternative Ideas 20%
• Uncertain Variable 20%
• Variable Model 20%
• Popular 20%
• Deep Reinforcement Learning (deep RL) 20%
• Deep Reinforcement Learning Control 20%

Engineering

• Recurrent 100%
• Reinforcement Learning 100%
• Predictive Control Model 66%
• Observables 66%
• Accurate Prediction 33%
• Performance Improvement 33%
• Observability 33%
• Learning Control 33%
• Rule-Based Control 33%
• Heat Pump 33%
• Markov Decision Process 33%

Chemical Engineering

• Reinforcement Learning 100%
• Predictive Control Model 66%