Multi-Timescale Reward-Based DRL Energy Management for Regenerative Braking Energy Storage System

Keyphrases

• Energy Management 100%
• Energy Storage System 100%
• Multiple Time Scales 100%
• Reward-based 100%
• Deep Reinforcement Learning (deep RL) 100%
• Regenerative Braking Energy 100%
• Rail Transportation Power Systems 50%
• Management Objectives 33%
• Learning-based 33%
• Optimization Methods 16%
• Rule-based 16%
• System Controller 16%
• Operation Conditions 16%
• Traditional Rule 16%
• Optimal Strategy 16%
• System Integration 16%
• Load Data 16%
• Markov Decision Process 16%
• Rule Optimization 16%
• Power Energy 16%
• Field Load 16%
• Reward Function 16%
• Power Demand 16%
• Reinforcement Learning Algorithm 16%
• Shaving 16%
• Multi-objective Energy Management 16%
• Deep Q-learning Algorithm 16%
• Deep Q-learning 16%
• Energy Management Problem 16%
• OPAL-RT 16%
• Traditional Models 16%
• Regenerative Braking Energy Utilization 16%
• Multiple Energy 16%
• Model-free Deep Reinforcement Learning 16%

Engineering

• Energy Storage System 100%
• Regenerative Braking 100%
• Energy Management 100%
• Reinforcement Learning 100%
• Railway 42%
• Power Engineering 42%
• Experimental Platform 14%
• Multistage 14%
• Q-Learning 14%
• Learning Algorithm 14%
• Markov Decision Process 14%