电工技术学报  2019, Vol. 34 Issue (19): 4135-4142    DOI: 10.19595/j.cnki.1000-6753.tces.181385
电力系统 |
基于深度置信网络的电力负荷识别
徐春华1, 陈克绪2, 马建2, 刘佳翰1, 吴建华1
1. 南昌大学信息工程学院 南昌 330031;
2. 国网江西省电力有限公司电力科学研究院 南昌 330096
Recognition of Power Loads Based on Deep Belief Network
Xu Chunhua1, Chen Kexu2, Ma Jian2, Liu Jiahan1, Wu Jianhua1
1. School of Information Engineering Nanchang University Nanchang 330031 China;
2. Institute of Electric Power Science State Grid Jiangxi Electric Power Co. Ltd Nanchang 330096 China
全文: PDF (1542 KB)  
输出: BibTeX | EndNote (RIS)      
摘要 针对目前电力负荷自动识别中存在的人工选择特征困难的问题并且为了进一步提高识别精度,提出一种基于深度置信网络(DBN)的电力负荷识别方法。DBN是一种深度神经网络架构,在图像识别、语音识别以及电能质量扰动识别等领域有着成功的应用。DBN由多个受限玻耳兹曼机(RBMs)和一层后向传播神经网络组成。使用对比散度算法,首先对第一个RBM进行充分训练,获取初始特征;然后将这些初始特征值作为训练数据训练下一个RBM以获取高级特征,以此类推;最后,通过反向传播算法采用监督方式微调整个DBN。实验结果表明所提方法在8种电力负荷类型的识别上有很好的效果,平均识别率超过98%。
服务
把本文推荐给朋友
加入我的书架
加入引用管理器
E-mail Alert
RSS
作者相关文章
徐春华
陈克绪
马建
刘佳翰
吴建华
关键词 电力负荷深度学习受限玻耳兹曼机深度置信网络识别对比散度    
Abstract:Aiming at the difficulty of manual feature selection in current automatic recognition of power loads (PLs) and to further heighten the identification accuracy, a deep belief network (DBN) based recognition method is proposed in this paper. DBN is a framework of deep neural network and has found a wide use in image recognition, speech recognition, power quality disturbances recognition, etc. A DBN consists of several restricted Boltzmann machines (RBMs) and one layer of back-propagation neural network. By using the contrastive divergence algorithm, the first RBM is fully trained with the training data to obtain initial features, then the next RBM is trained with the initial features as training data, etc. Finally, the whole DBN is fine-tuned in a manner of supervised training by back-propagation. Experimental results demonstrate that the proposed approach has a good performance on the recognition of eight types of PLs with an average accuracy over 98%.
Key wordsPower loads    deep learning    restricted Boltzmann machines    deep belief network    recognition    contrastive divergence   
收稿日期: 2018-08-17      出版日期: 2019-10-12
PACS: TM711  
基金资助:国家自然科学基金项目(61662047)和国网江西省电力有限公司科技项目(521820180014)资助
通讯作者: 吴建华 男,1956年生,教授,博士生导师,研究方向为数字图像处理、电能质量分析、模式识别。E-mail:jhwu@ncu.edu.cn   
作者简介: 徐春华 男,1989年生,硕士研究生,研究方向为电力负荷类型识别、深度学习。E-mail:chunhua_ly@163.com
引用本文:   
徐春华, 陈克绪, 马建, 刘佳翰, 吴建华. 基于深度置信网络的电力负荷识别[J]. 电工技术学报, 2019, 34(19): 4135-4142. Xu Chunhua, Chen Kexu, Ma Jian, Liu Jiahan, Wu Jianhua. Recognition of Power Loads Based on Deep Belief Network. Transactions of China Electrotechnical Society, 2019, 34(19): 4135-4142.
链接本文:  
https://dgjsxb.ces-transaction.com/CN/10.19595/j.cnki.1000-6753.tces.181385          https://dgjsxb.ces-transaction.com/CN/Y2019/V34/I19/4135