Digital twin based reinforcement learning for extracting network structures and load patterns in planning and operation of distribution systems

Keyphrases

• Distribution System 100%
• Network Structure 100%
• Structure Pattern 100%
• Reinforcement Learning 100%
• Digital Twin 100%
• Network Load 100%
• Load Pattern 100%
• System Transition 100%
• Convolutional Neural Network 66%
• System Dynamics 33%
• Pattern Recognition 33%
• Iterative Algorithm 33%
• Power Consumption 33%
• Computation Time 33%
• Optimization Problem 33%
• Batch Mode 33%
• Power Loss 33%
• Operational Strategy 33%
• Planning Decision 33%
• Decision Support 33%
• Recognition Ability 33%
• Investment Cost 33%
• Distribution Network 33%
• Combination Pattern 33%
• Planning Strategy 33%
• Network Configuration 33%
• Training Model 33%
• Systems Change 33%
• Electric Vehicle 33%
• Low-carbon Technology 33%
• Benchmark Optimization 33%
• Optimal Decision 33%
• System Operator 33%
• Last Mile 33%
• Heat Electrification 33%
• Heat Storage 33%
• Distribution System Operator 33%
• Future Energy Scenarios 33%
• High-dimensional Input 33%
• Operational Decisions 33%
• Power System State 33%
• Fitted Q-iteration 33%
• Low-voltage Distribution Network 33%
• Legacy Assets 33%
• Scalable Deployment 33%
• Load Loss 33%

Engineering

• Distribution System 100%
• Reinforcement Learning 100%
• Electric Power Distribution 66%
• Convolutional Neural Network 66%
• Power Engineering 33%
• Electric Power Utilization 33%
• Computational Time 33%
• Metrics 33%
• Pattern Recognition 33%
• Future Energy 33%
• Investment Cost 33%
• Iteration Algorithm 33%
• Electric Vehicle 33%
• Power Loss 33%
• Responsiveness 33%
• Heat Storage 33%
• Solving Optimization Problem 33%
• System Operator 33%
• System State 33%
• Distribution System Operator 33%
• Optimal Decision 33%
• Dimensional Input 33%
• Energy Scenario 33%
• Electric Load Loss 33%

Chemical Engineering

• Reinforcement Learning 100%
• Neural Network 66%