摘要
针对电力市场中存在潜在危害性交易行为难以被识别、管控的问题,提出基于云模型与模糊Petri网的电力市场潜在危害行为识别方法,在交易过程中对潜在危害行为进行实时预警。将云模型与模糊Petri网应用于模式识别中,有效解决了潜在危害行为难以量化识别的问题,也克服了一般识别方法不易理解的缺陷。首先,设计潜在危害行为识别的一般框架与步骤;其次,阐述通过逆向云发生器与根据专家描述的两种特征云的构建方法;同时,设计潜在危害行为识别的模糊Petri网网络结构与识别算法;最后,以江西省月度竞价交易市场为背景,采用真实交易数据进行算例分析。算例分析结果表明:所设计的方法给出的结果与专家分析得出的结果以及实际的数据高度一致,验证了方法的有效性。
To solve the problem of identification and control of potential harmful transactions in electricity market,a method for identifying potential harmful behaviors based on the cloud model and the fuzzy Petri net is proposed to realize real-time warning of potential hazards in the transaction process.The cloud model and the fuzzy Petri net are applied to pattern recognition,which effectively solves the problem that the potential hazard behavior is difficult to quantify,and overcomes the defects that the general recognition method is not easy to understand.Firstly,ageneral framework and procedure for identifying potential harmful behaviors is designed.Secondly,two methods of building feature cloud based on reverse cloud generator and expert description are described.Meanwhile,a fuzzy Petri net structure and recognition algorithm for potential harmful behaviors is designed.Finally,an example is made and analyzed by using the real transaction data of monthly auction market in Jiangxi Province,which shows the results given by the proposed method are highly consistent with the results obtained by artificial analysis and actual data.
作者
刘敦楠
张潜
李霄彤
顾宇桂
谭忠富
LIU Dunnan;ZHANG Qian;LI Xiaotong;GU Yugui;TAN Zhongfu(State Key Laboratory of Alternate Electrical Power System with Renewable Energy Sources (North China Electric Power University),Beijing 102206,China;Beijing Electric Power Trading Center Co.Ltd.,Beijing 100031,China)
出处
《电力系统自动化》
EI
CSCD
北大核心
2019年第2期25-33,共9页
Automation of Electric Power Systems
基金
国家电网公司总部科技项目"我国电力市场运营成效动态评估监测体系及信用评价体系研究"~~
关键词
电力市场
市场监管
潜在危害行为
模式识别
云模型
模糊PETRI网
electricity market
market supervision
potential harmful behavior
pattern identification
cloud model
fuzzy Petri net
