Page 18 - 电力与能源2024年第三期
P. 18
第 45 卷 第 3 期
电力与能源
292 2024 年 6 月
DOI:10.11973/dlyny202403003
基于组合赋权—云模型的电力市场
主体信用风险评估
栾晶晶,谢敬东
(上海电力大学 能源电力科创中心,上海 200090)
摘 要:在电力市场中,市场主体信用风险的存在会对市场秩序、市场参与者利益和电力供需平衡等方面产生
重大影响。提出了一种基于组合赋权-云模型的电力市场主体信用风险评估方法。该方法基于“结构—行为
—绩效”框架设计了包含 3 个一级指标 14 个二级指标的主体信用风险评估指标体系,同时以博弈论为基础构
建了属性层次模型与反熵权法的指标组合赋权法,并提出了基于云理论的电力市场主体信用风险评估模型。
试验结果表明,该方法能够全面、科学、准确地评估电力市场主体的信用风险,为电力市场的稳定发展提供有
效支撑。
关键词:电力市场;主体信用风险;组合赋权;云模型;风险评估
作者简介:栾晶晶(1986—),女,硕士,主要研究方向为电力市场、院校管理。
中图分类号:TM73 文献标志码:B 文章编号:2095-1256(2024)03-0292-08
Credit Risk Assessment of Power Market Entities Based on
Combined Weighting-Cloud Model
LUAN Jingjing,XIE Jingdong
(Energy and Power Science and Technology Innovation Center,Shanghai University of Electric Power,Shanghai 200090,
China)
Abstract:In the power market, the presence of credit risk among market entities can have significant impacts on
market order, interests of market participants, and the balance of power supply and demand. This paper proposes
a method for assessing the credit risk of power market entities based on combined weighting-cloud model. The
study designs a credit risk assessment index system for entities, consisting of 3 primary indicators and 14
secondary indicators, based on the "structure-behavior-performance" framework. Additionally, based on game
theory, a method of index combination weighting based on attribute hierarchy model and anti-entropy method is
constructed, and a credit risk assessment model for power market entities based on cloud theory is proposed.
Experimental results demonstrate that this method can comprehensively, scientifically, and accurately assess the
credit risk of power market entities, providing effective support for the stable development of the power market.
Key words:power market,entity credit risk,combined weighting,cloud model,risk assessment
随着我国电力市场化改革的加速,各项改革 出了一种售电公司信用评价的方法,实现了评价
措施有序推进,在多个领域取得了显著成效和进 指标的降维,解决了模型对象模糊性问题。文献
展。市场履约是保障电力市场交易的关键环节, [5]介绍了电力市场信用风险对电力市场的影响,
而 市 场 违 约 是 电 力 市 场 运 营 中 常 见 的 风 险 总结了国际典型电力市场信用评级管理经验。文
问题 [1-3] 。 献[6-7]以电力用户历史缴费记录信息为基础,使
目前,国内外在电力市场主体信用风险评估 用逼近理想解排序法与主成分分析法,提出了一
与管控领域已经取得了一定的进展。文献[4]提 种电力用户的信用评价模型。文献[8]基于偏差