Virtual power plant economic dispatching method in energy internet based on deep reinforcement learning

An energy Internet and reinforcement learning technology, applied in the field of energy distribution of virtual power plants, can solve problems such as high computational complexity, large communication load and delay, and poor reliability of data transmission, so as to preserve privacy, reduce communication load, and reduce computational complexity degree of effect

Active Publication Date: 2020-06-05
STATE GRID HEILONGJIANG ELECTRIC POWER COMPANY +1
View PDF6 Cites 13 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to solve the problems of high computational complexity, large communication load and delay, and poor reliability of data transmission in the existing methods, and propose a virtual power plant economic dispatch in the energy Internet based on deep reinforcement learning method

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Virtual power plant economic dispatching method in energy internet based on deep reinforcement learning
  • Virtual power plant economic dispatching method in energy internet based on deep reinforcement learning
  • Virtual power plant economic dispatching method in energy internet based on deep reinforcement learning

Examples

Experimental program
Comparison scheme
Effect test

specific Embodiment approach 1

[0024] Specific implementation mode 1: The economic scheduling method of virtual power plants in the energy Internet based on deep reinforcement learning described in this implementation mode, the method includes the following steps:

[0025] Step 1. For any area i, use the industrial side server and power supply side server of area i to collect the power generation side and user side information from area i, i=1,2,...,I, I is the total number of areas;

[0026] Use the information collected in each region to train the actor-critic network of the cloud server of the VPP operator, so as to obtain the actor-critic network trained by using the information of each region;

[0027] Step 2. Deploy the trained actor-critic network on the edge nodes of the corresponding regions;

[0028] Step 3: The industrial-side servers and power-side servers in each region collect information from the power generation side and the user side in real time, and input the collected information into th...

specific Embodiment approach 2

[0029] Specific embodiment 2: The difference between this embodiment and specific embodiment 1 is that in the first step, the actor-critic network of the VPP operator cloud server is trained using the information collected in each area, and an asynchronous method is adopted. , running 8 threads in parallel at the same time.

specific Embodiment approach 3

[0030] Embodiment 3: The difference between this embodiment and Embodiment 1 is that the objective function of the actor-critic network is:

[0031]

[0032] Where: C is the total operating cost of area i, is the initial depreciation cost of photovoltaic investment in region i at time slot k, k=0,1,...,K (considering 24 hours in VPP, K is equal to 23), is the photovoltaic operation and maintenance cost of area i in time slot k, is the initial depreciated cost of wind turbines in region i at time slot k, is the wind turbine operation and maintenance cost in region i at time slot k, is the initial depreciation cost of the micro gas turbine in region i at time slot k, is the operation and maintenance cost of the micro gas turbine in region i at time slot k, is the environmental cost of micro gas turbines in region i at time slot k, is the cost consumed by the micro gas turbine in area i at time slot k, λ is the compensation coefficient, is the controllable load ...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention discloses a virtual power plant economic dispatching method in an energy internet based on deep reinforcement learning, and belongs to the technical field of energy distribution of virtual power plants. According to the invention, the problems of large communication load and delay, high calculation complexity and poor data transmission reliability in the existing method are solved. The invention provides a distributed power generation economic dispatching structure using a three-layer system structure based on edge computing, in which a first layer and a second layer are edge computing layers, and a third layer is a cloud computing layer. According to the proposed three-layer edge computing architecture, the computing complexity of processing the training task at the centralnode is reduced, and the communication load between the VPP operator and the DG is further reduced, so that the response time of the industrial user is also reduced, the privacy of the industrial useris also retained, and the reliability of data transmission is improved. The method can be applied to energy distribution of the virtual power plant.

Description

technical field [0001] The invention belongs to the technical field of energy distribution of virtual power plants, and in particular relates to an economic scheduling method of virtual power plants in the energy Internet based on deep reinforcement learning. Background technique [0002] With the access of large-scale distributed power generation in the energy Internet, due to the limitation of geographical conditions, the traditional micro-grid has certain limitations, which hinder the effective utilization of large-scale distributed power generation in multiple regions, and the power cuts are very frequent. Due to the mismatch between the construction scale of renewable energy stations and the demand of local loads, the capacity of renewable energy is limited, resulting in a certain amount of power cuts in areas where wind power stations and photovoltaic power stations are concentrated. Compared with microgrids, VPP has a larger energy load channel, which can better match...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Applications(China)
IPC IPC(8): G06Q10/06G06Q50/06
CPCG06Q10/06312G06Q50/06Y04S10/50Y02E40/70
Inventor 孙迪王宁关心林霖
Owner STATE GRID HEILONGJIANG ELECTRIC POWER COMPANY
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap