Micro-grid energy management method based on Rainbow deep Q network

An energy management and microgrid technology, applied in AC network circuits, neural learning methods, AC network load balancing, etc. Difficulty effects

CN114742453APending Publication Date: 2022-07-12JIANGSU UNIV +2
0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Publication Date
2022-07-12

Smart Images

  • Figure 1
    Figure 1
  • Figure 2
    Figure 2
  • Figure 3
    Figure 3
Patent Text Reader

Abstract

The invention discloses a micro-grid energy management method based on a Rainbow deep Q network, and the method comprises the steps: taking an electric vehicle and energy storage equipment in a micro-grid of a residential area as controlled carriers; load demand, photovoltaic and wind power output, and charge state prediction data of an electric vehicle and energy storage equipment of a micro-grid in 24 hours of a day are used as state space input of a Rainbow deep Q network, and charge and discharge actions of the electric vehicle and the energy storage equipment are learned and executed through an intelligent agent. Therefore, the purpose of lowest one-day operation cost of the micro-grid on the premise of ensuring stable power and safe service life of equipment is achieved. According to the method, the problem that the micro-grid energy management model is difficult to model due to randomness and intermittency of renewable energy sources can be solved, and the method has rapid convergence after training and also has excellent generalization ability when coping with a new micro-grid energy management model; therefore, the defects of low micro-grid energy utilization rate and relatively high operation cost caused by uncertainty of renewable energy sources and diversity of energy flow and loads can be effectively overcome.
Need to check novelty before this filing date? Find Prior Art

Description

technical field

[0001] The invention relates to the technical field of electric power engineering, in particular to the field of microgrid operation control and energy management. Background technique

[0002] In recent years, due to the shortage of global coal inventories, the power generation of power plants has continued to show negative growth, resulting in global power shortages. As a new type of energy Internet coupled with multiple energy systems and energy storage devices, microgrid can not only access clean new energy such as wind power and photovoltaic power to reduce the power purchase of the grid, but also utilize distributed resources and various Energy management means to improve energy utilization, reduce energy and economic waste. When dealing with power shortages, it can play a positive role in increasing revenue and reducing expenditure. However, it is precisely because of the multi-energy coupling characteristics of microgrids, the uncertainty of renewab...

Examples

Embodiment Construction

[0058] The technical solutions of the embodiments of the present invention will be explained and described below, but the following embodiments are only preferred embodiments of the present invention, not all. Based on the examples in the implementation manner, other examples obtained by those skilled in the art without creative work all belong to the protection scope of the present invention.

[0059] The invention relates to an energy management method, which describes the energy management problem of a microgrid as a Markov decision process under a reinforcement learning framework, and learns an optimal control strategy after continuously exchanging state information with the environment.

[0060] The purpose of the present invention is to input the real-time state information of each scheduling period into the neural network in the scheduling stage, and the neural network evaluates each action and selects the optimal action, and executes the electric vehicle and the energy ...