Reinforcement Learning Enabled Real-Time Energy Dispatch of Energy Systems with Flexible Operational Resources
- Resource Type
- Conference
- Authors
- Gong, Liwu; Huang, Yuehua; Zhang, Wei; Gu, Yixing; Chen, Chao
- Source
- 2023 International Conference on Ambient Intelligence, Knowledge Informatics and Industrial Electronics (AIKIIE) Ambient Intelligence, Knowledge Informatics and Industrial Electronics (AIKIIE), 2023 International Conference on. :1-6 Nov, 2023
- Subject
- Communication, Networking and Broadcast Technologies
Computing and Processing
Engineered Materials, Dielectrics and Plasmas
Engineering Profession
General Topics for Engineers
Deep learning
Job shop scheduling
Reinforcement learning
Control systems
Real-time systems
Generators
Stability analysis
Multi-Energy Systems
Reinforcement Learning
Optimization
Day-Ahead Energy Planning
- Language
This paper developed a reinforcement learning- enabled real-time energy dispatch solution for energy systems with flexible operational resources. The detailed process of the proposed solution is formulated and implemented. The deep reinforcement learning method is introduced to solve the action strategy of the energy storage system and the controllable generator. Specifially, this work developed a reinforcement learning-enabled real-time energy dispatch solution for energy systems considering the presence of flexible operational resources. For the real-time scheduling phase, the timing decision problem of real-time scheduling is designed to maintain the stability of the trading curve. The deep reinforcement learning algorithm is designed and adopted to identify the action strategy of the energy storage system and the controllable generator. The proposed solution is evaluated through a case study via simulations and the numerical results confirmed the effectiveness of the proposed real-time energy dispatch solution for multi-energy systems.