摘要
为解决电力系统暂态过电压风险评估中输入特征集构建合理性不足、强相关性差等问题,提出一种考虑新能源多场站短路比MRSCR(multiple renewable energy stations short circuit ratio)的暂态过电压风险评估方法。首先,通过分析暂态过电压数学模型,发现MRSCR与暂态过电压呈负相关性;然后,综合考虑MRSCR与其他影响系统暂态过电压的关键因素,构建多维输入特征集;最后,通过卷积神经网络建立输入特征与暂态过电压的高维映射,实现系统暂态过电压风险的快速、准确评估,并通过算例分析验证了所提方法的有效性、可行性。
To solve the problems in the transient overvoltage risk assessment of power system such as insufficient rationality and poor strong correlation of input feature set construction,a transient overvoltage risk assessment method considering the multiple renewable energy stations short circuit ratio(MRSCR)is proposed.Firstly,by analyzing the mathematical model of transient overvoltage,it is found that MRSCR is negatively correlated with transient overvoltage.Then,considering MRSCR and other key factors affecting the transient overvoltage of the system,a multi-dimensional input feature set is constructed.Finally,the high-dimensional mapping between input features and transient overvoltage is established by convolutional neural networks to achieve a rapid and accurate assessment of system transient overvoltage risk.The effectiveness and feasibility of the proposed method are verified by the analysis of an example.
作者
付小标
崔健
王长江
宋晓喆
孙勇
FU Xiaobiao;CUI Jian;WANG Changjiang;SONG Xiaozhe;SUN Yong(State Grid Jilin Electric Power Supply Company,Changchun 130021,China;Key Laboratory of Modern Power System Simulation and Control&Renewable Energy Technology of Ministry of Education(Northeast Electric Power University),Jilin 132012,China)
出处
《电力系统及其自动化学报》
CSCD
北大核心
2024年第10期79-88,共10页
Proceedings of the CSU-EPSA
基金
国网吉林省电力有限公司科技项目(SGJL0000DKJS2200369)。
关键词
高比例新能源电力系统
新能源多场站短路比
输入特征构建
暂态过电压风险评估
卷积神经网络
high penetration new energy power system
multiple renewable energy stations short circuit ratio(MRSCR)
input feature construction
transient overvoltage risk assessment
convolutional neural networks
