上海交通大学学报(自然版) ›› 2018, Vol. 52 ›› Issue (10): 1206-1219.doi: 10.16183/j.cnki.jsjtu.2018.10.008

• 学报(中文) • 上一篇    下一篇

“人工智能+”时代下的智能电网预测分析

吴倩红,韩蓓,冯琳,李国杰,江秀臣   

  1. 上海交通大学 电力传输与功率变换控制教育部重点实验室, 上海 200240
  • 通讯作者: 韩蓓, 女, 讲师, E-mail: han_bei@sjtu.edu.cn.
  • 作者简介:吴倩红(1991-), 女, 山西省临汾市人, 博士生, 主要研究方向为大数据在智能电网中的应用.

“AI+” Based Smart Grid Prediction Analysis

WU Qianhong,HAN Bei,FENG Lin,LI Guojie,JIANG Xiuchen   

  1. Key Laboratory of Control of Power Transmission and Conversion of Ministry of Education, Shanghai Jiao Tong University, Shanghai 200240, China

摘要: 智能电网预测分析是保证智能电网经济、安全运行的基础.借助人工智能的突破性技术以及智能电网的大数据环境,实现基于人工智能的智能电网预测分析对电力系统发展具有重大意义,为此提出了“人工智能+”预测.首先介绍了人工智能与智能电网预测分析的背景及所涉及的问题;然后根据应用的不同侧重点,展开人工智能在新能源预测、负荷预测、静态电压稳定预测及其相关预防性措施三个方面的研究综述及研究展望,并对预测中所涉及的其他相关技术(数据样本产生、不平衡样本、特征提取)进行了总结;最后对人工智能局限性及发展进行了讨论,并提出了一些建议与设想.

关键词: 智能电网, “人工智能+”预测, 数据样本产生, 不平衡样本, 特征提取

Abstract: Smart grid prediction analysis (SGPA) is the basis to ensure the economic and safe operation of power system. With the breakthrough of AI and the big data environment for smart grid, AI based SGPA is significant for the development of power system, thus this paper proposes “AI+” prediction. Firstly, the background of AI and SGPA and some related issues are introduced. Then based on different emphasis, the AI applications in renewable energy prediction, load prediction, steady state voltage stability prediction and related preventive measures are reviewed and future direction of the research content is prospected. In addition, other related technologies involved in prediction (sample generation, imbalanced data and features extraction) are summarized. Finally, limitations and future developments about AI are discussed and suggestions and ideas are proposed.

Key words: smart grid, “AI+” prediction, sample generation, imbalanced data, features extraction

中图分类号: