Real-Time Holding Control for Transfer Synchronization via Robust Multiagent Reinforcement Learning

Keyphrases

• Optimization Control 100%
• Multi-agent Reinforcement Learning 100%
• Holding Control 100%
• Rolling Horizon Optimization 100%
• Transfer Synchronization 100%
• Reinforcement Learning Approach 100%
• Average Waiting Time 75%
• Network Applications 50%
• Transit Network 50%
• Reinforcement Learning Algorithm 50%
• System Uncertainty 50%
• Twin Cities 25%
• Minnesota 25%
• Learning Objectives 25%
• Specific Learning 25%
• Real-time Networks 25%
• Computation Time 25%
• Passenger Vehicle 25%
• Synchronization Problem 25%
• Computational Efficiency 25%
• Computational Results 25%
• Computational Solutions 25%
• Learning Procedure 25%
• Daily Operation 25%
• Joint Action 25%
• Online Computation 25%
• Control Rules 25%
• Rule-based Control 25%
• Policy Performance 25%
• Passenger Demand 25%
• Robust Policy 25%
• Deep Deterministic Policy Gradient Algorithm 25%
• Actor-critic Methods 25%
• Disrupted Environments 25%
• Robust Deep Reinforcement Learning 25%
• Short-term Uncertainty 25%
• Stochastic Scenarios 25%
• Max-min 25%
• Reinforcement Learning Framework 25%
• Deterministic Scenario 25%

Engineering

• Multiagents 100%
• Holding Time 100%
• Reinforcement Learning 100%
• Learning Approach 50%
• Joints (Structural Components) 12%
• Collected Data 12%
• Relative Performance 12%
• Computational Efficiency 12%
• Computation Time 12%
• Rule-Based Control 12%
• Computational Solution 12%
• Daily Operation 12%
• Synchronization Problem 12%
• Approximators 12%

Computer Science

• Reinforcement Learning 100%
• multi-agent 100%
• Learning Approach 66%
• Learning Framework 16%
• Theoretical Lower Bound 16%
• Online Computation 16%
• Deep Reinforcement Learning 16%
• Collected Data 16%
• Computation Time 16%
• Computational Efficiency 16%
• Computational Solution 16%
• Relative Performance 16%
• Multi-Agent Reinforcement Learning 16%