The integration of renewable energy sources and carbon capture system in cyber physical energy system is the basis for coping with the challenge of global warming. It is very important to give full play to the role of carbon capture equipment to maximize carbon emission reduction. First of all, we propose a joint dispatch model for wind power, energy storage, and thermal power units, which may minimize the operation costs of system. Then an optimal dispatch model for wind power, energy storage, carbon capture, and thermal power units is established. In this model, the optimal operation cost of the economic dispatch with no carbon capture equipment is considered as one of the constraints, and the optimization goal is to minimize the carbon emissions of the economic dispatch. Due to the fact that the dispatch problem can be defined as a Markov decision process, deep reinforcement learning model is employed to solve the optimal dispatch problem in cyber physical energy system. The experimental results indicate that the proposed scheduling model can effectively reduce operation costs and carbon emissions.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Deep Reinforcement Learning Based Economic Dispatch with Cost Constraint in Cyber Physical Energy System

  • Xue Chen,
  • Chen Cui,
  • Ning Wang,
  • Zhaogong Zhang,
  • Jinghong He,
  • Xin Guan

摘要

The integration of renewable energy sources and carbon capture system in cyber physical energy system is the basis for coping with the challenge of global warming. It is very important to give full play to the role of carbon capture equipment to maximize carbon emission reduction. First of all, we propose a joint dispatch model for wind power, energy storage, and thermal power units, which may minimize the operation costs of system. Then an optimal dispatch model for wind power, energy storage, carbon capture, and thermal power units is established. In this model, the optimal operation cost of the economic dispatch with no carbon capture equipment is considered as one of the constraints, and the optimization goal is to minimize the carbon emissions of the economic dispatch. Due to the fact that the dispatch problem can be defined as a Markov decision process, deep reinforcement learning model is employed to solve the optimal dispatch problem in cyber physical energy system. The experimental results indicate that the proposed scheduling model can effectively reduce operation costs and carbon emissions.