摘要
电力系统暂态稳定性的破坏可以对电力系统的安全稳定运行产生严重冲击,准确、快速地暂稳评估方法能够提高电力系统的安全防御能力。极限学习机由于其速度快、泛化性能好被应用到电力系统暂态稳定评估中。为了提高极限学习机的评估性能,利用基于差分进化算法的优化方法和序列浮动后向特征选择算法对极限学习机暂态稳定评估性能进行提升。首先对输入特征通过主元分析降维并利用序列浮动后向算法进行特征选择,再将最优特征集输入差分进化极限学习机进行暂态稳定评估,最后在新英格兰10机39节点系统中进行验证分析,结果表明,所提模型与其他极限学习机模型相比,大大提升了其在暂态稳定分类评估中的性能。
A damage in power system transient stability can cause serious accident of the power network;thus,accurate and fast assessment of the transient stability is of great significance for the system safety and stable operation.A method based on extreme learning machine was developed and optimized in the differential evolution algorithm(DE-ELM)and the sequence floating backward feature selection(SFBS).First,the dimension of input features was reduced by principal component analysis,and feature selection was performed by sequential floating backward algorithm.Second,an optimal feature set was input into the DE limit learning machine for the assessment.Finally,calculation was verified in the New England 10-machine 39-node system.The results show that the proposed model was superior in the classification and assessment of transient stability to those of other limit learning machine models.
作者
李向伟
刘思言
高昆仑
LI Xiang-wei;LIU Si-yan;GAO Kun-lun(School of Electric College,North China Electric Power University,Beijing 102206,China;Global Energy Interconnection Research Institute,Beijing 102209,China)
出处
《科学技术与工程》
北大核心
2020年第1期213-217,共5页
Science Technology and Engineering
基金
国家电网公司科技项目(SGGR0000JSJS1800569)
关键词
极限学习机
差分进化算法
序列浮动后向特征选择
暂态稳定评估
extreme learning machine
differential evolution algorithm
sequence floating backward feature selection
transient stability assessment
作者简介
第一作者:李向伟(1995-),男,汉族,河北唐山人,硕士研究生。研究方向:电力系统人工智能。E-mail:469942856@qq.com。通信作者:高昆仑(1972-),男,汉族,吉林人,博士,高级工程师。研究方向:电力系统人工智能。E-mail:GKL@GEIRI.SGCC.com。
