摘要
随着园区微电网型虚拟电厂(virtual power plant,VPP)的普及,多个虚拟电厂之间的分布式交易逐渐成为研究热点。VPP之间的直接交易中引入区块链技术提高了交易的可追溯性和安全性,降低了中间环节的交易成本,但电网运行的物理约束需充分考虑,每笔交易的潮流效果需在交易中予以反映。为此,提出基于机器学习的潮流分布识别方法将阻塞信息自动接入区块链系统,利用虚拟电厂中的储能系统进行阻塞管理。首先,设计可报价调整的三阶段交易流程以保证无法随时更改报价的区块链成功匹配更多的交易;然后,提出采用人工智能技术进行潮流判断的智能化方案,实现安全校核自动化;最后,提出利用电动汽车储能特性进行阻塞管理的思路。算例仿真的结果验证了交易方案可以降低交易成本,基于机器学习的潮流分布识别准确率可以满足交易结果安全校核的需求。
With the popularization of park microgrid-type virtual power plant(VPP),distributed transactions between multiple virtual power plants have gradually become a research hotspot.The introduction of blockchain technology in the direct transactions between VPPs improves the traceability and security of the transactions and reduces the transaction costs of the intermediate links,but the removal of the centralized management organization makes the physical constraints of the power grid operation unaccountable,but the physical constraints of grid operation need to be fully considered,and the power flow effect of each transaction needs to be reflected in the transaction.To this end,a power flow distribution identification method based on machine learning is proposed to automatically connect the blocking information to the blockchain system,and use the energy storage system in the virtual power plant for blocking management.Firstly,a three-stage transaction process with adjustable quotes is designed to ensure that the blockchain,which cannot change the quotes at any time,successfully matches more transactions;secondly,an intelligent scheme of tidal current judgment using artificial intelligence technology is proposed to realize the automation of safety calibration;and finally,the blocking problem is solved by using the energy storage characteristics of electric vehicles.The results of the algorithmic simulation verify that the transaction scheme can reduce the transaction cost,while the accuracy of machine learning-based power flow distribution identification can meet the demand for security calibration of transaction results.
作者
高赐威
陈彦
石坤
王佳颖
肖涛
GAO Ciwei;CHEN Yan;SHI Kun;WANG Jiaying;XIAO Tao(School of Electrical Engineering Southeast University,Nanjing 210018,China;College of Software Engineering Southeast University,Suzhou 215125,China;China Electric Power Research Institute,Beijing 100089,China;State Grid Zhejiang Electric Power Company,Ltd.Marketing Service Center,Hangzhou 311100,China)
出处
《供用电》
北大核心
2024年第8期112-119,共8页
Distribution & Utilization
基金
国家电网有限公司总部项目“面向新兴可调资源聚合的精准响应技术与市场化交易研究及应用项目”(5400-202319244A-1-1-ZN)。
关键词
虚拟电厂
区块链
电力交易
人工智能
潮流
virtual power plant
blockchain
power transaction
artificial intelligence
power flow
