摘要
为准确识别违规的分布式光伏扩容骗补用户,提出一种基于改进深度极限学习机的光伏扩容用户识别方法。首先利用同地区光伏发电出力具有相似性的特点,通过余弦相似度对参考电站和待测站点进行预处理;然后应用麻雀搜索算法SSA(sparrow search algorithm)对深度极限学习机DELM(deep extreme learning machine)的权值参数优化,用预处理的数据集训练SSA-DELM拟合模型,并根据光伏扩容的特性计算扩容系数。实验结果验证了所提方法对分布式光伏违规扩容用户识别的有效性。
To accurately identify the distributed photovoltaic(PV)capacity expansion fraud users,an identification method for PV capacity expansion users based on an improved deep extreme learning machine(DELM)is proposed.First,with the consideration of the similarity of PV generation in the same region,the reference power station and the site to be tested are pre-processed by cosine similarity.Second,the sparrow search algorithm(SSA)is used to optimize the weight parameters of DELM,and the pre-processed data set is imported into the SSA-DELM fitting model.Finally,the expansion coefficients are calculated according to the characteristics of PV capacity expansion.Experimental results validate the effectiveness of the proposed method for the identification of distributed PV non-compliant capacity expansion users.
作者
汤渊
吴裕宙
苏盛
刘韵艺
王耀龙
TANG Yuan;WU Yuzhou;SU Sheng;LIU Yunyi;WANG Yaolong(Dongguan Power Supply Bureau,Guangdong Power Grid Co.,Ltd.,Dongguan 523008,China;State Key Laboratory of Disaster Prevention&Reduction for Power Grid,Changsha University of Science and Technology,Changsha 410004,China)
出处
《电力系统及其自动化学报》
CSCD
北大核心
2024年第5期59-68,共10页
Proceedings of the CSU-EPSA
基金
国家自然科学基金资助项目(51777015)
南方电网公司科技项目(031900KK52220039)。
关键词
分布式光伏
违规扩容
深度极限学习机
麻雀搜索算法
distributed photovoltaic(PV)
non-compliant capacity expansion
deep extreme learning machine(DELM)
sparrow search algorithm(SSA)
