文章摘要
刘博,柯德平,李鹏,等.低压配电台区分布式光伏发电功率辨识方法[J].电力系统自动化,2019,43(19):111-116. DOI: 10.7500/AEPS20180912009.
LIU Bo,KE Deping,LI Peng, et al.Identification Method of Distributed Photovoltaic Power in Low-voltage Distribution Networks[J].Automation of Electric Power Systems,2019,43(19):111-116. DOI: 10.7500/AEPS20180912009.
低压配电台区分布式光伏发电功率辨识方法
Identification Method of Distributed Photovoltaic Power in Low-voltage Distribution Networks
DOI:10.7500/AEPS20180912009
关键词: 光伏发电  神经网络  线性相关系数  低压配电台区
KeyWords: photovoltaic generation  neural network  linear correlation coefficient  low-voltage distribution network
上网日期:2019-05-08
基金项目:国家重点研发计划资助项目(2017YFB0902900)
作者单位E-mail
刘博 武汉大学电气与自动化学院, 湖北省武汉市 430072  
柯德平 武汉大学电气与自动化学院, 湖北省武汉市 430072 kedeping@whu.edu.cn 
李鹏 南方电网科学研究院有限责任公司, 广东省广州市 510663  
徐箭 武汉大学电气与自动化学院, 湖北省武汉市 430072  
白浩 南方电网科学研究院有限责任公司, 广东省广州市 510663  
于力 南方电网科学研究院有限责任公司, 广东省广州市 510663  
摘要:
      数量众多的小容量光伏电源接入低压配电台区,能够实现可再生能源发电的就地消纳,但通常因量测不足而无法掌握这些光伏电源的出力情况,不利于配电网的调度与控制。以台区为单位,提出了一种台区内总的光伏发电功率辨识方法。基本思想是以实测的光照强度数据为输入,通过训练后的神经网络映射得到台区光伏发电功率,而该神经网络训练的目标则是光照强度与台区负荷之间的低线性相关性。通过理想的数值算例与基于实际电网数据的算例验证了该方法的有效性和可行性。
Abstract:
      A large number of small-capacity photovoltaic(PV)generators are integrated into low-voltage distribution networks, which can facilitate the accommodation of local renewable energy. However, owing to inadequate measurement units, it is generally difficult to precisely grasp PV output power in low-voltage distribution networks, and this situation has no benefits for the dispatch and control of distribution networks. To address the issue, a method is proposed to identify total power output of PV generation installed in a low-voltage distribution network. Specifically, based on the actual measured data of solar radiation intensity, the PV power output is identified through training a neural network with the objective of minimizing the linear correlation coefficient between the solar radiation intensity and the network load power. By using both the pure numerical data and the actual data from Guangdong power grids of China, simulation results have validated the effectiveness and feasibility of the proposed method.
查看全文(Free!)   查看附录   查看/发表评论  下载PDF阅读器