摘要
电力工业是一个国家在能源领域中的基础性行业,如何让电力工业的快速发展的同时能保持健康化是一个非常重要的问题。对于电力负荷预测来讲,预测精度的提高,将会对经济效益最优化的制定电力调配计划、制定发电计划以及制定上网竞价计划等方面都具有重要的意义,能产生直接而重大的经济效益和社会效益。由于短期负荷预测周期短,因此对安排日开停机计划和发电计划具有重要的意义,短期符合预测精度的高低直接影响其所起的作用,短期预测是目前人们所研究的主要方向。国内外对短期负荷的研究均远远多于中长期预测。
另一方面,随着配电网的智能化发展,电网中会逐渐引入智能化单元,比如各种分布式电源、电动汽车充电站、充电桩以及智能用电小区等。这些智能单元的引入将会对电网负荷模式和负荷增长模式产生重大影响,使得常规预测方法往往不能满足精度要求。为解决这一情况,本文提出一种拆分建模的方法。即将配电网中的常规部分和智能部分拆分,分别建模并进行负荷预测,最终将各部分的预测结果相结合,就可以得到整个配电网的负荷预测结果。
本文使用人工神经网络法和时间序列法进行短期负荷预测研究;智能配电网方面,重点研究了风力发电和太阳能光伏发电,对其功率输出预测进行研究,并对电动汽车充电站的负荷模型提出了展望。
The power industry is the foundation of the energy industries field. How to get the rapid development and the healthy of power industry at the same time is becoming a important problem. For the power load forecasting theory and technology, it is of great significance for optimizing the power generation plan, making the power deployment plan and making electricity price biding plan when the power load forecasting accuracy improved, it also have direct and significant economic and social benefits. Because the period of short-term power load forecasting is very short, so it is of great significant for arranging the day off on plan and making the power generation plan. The role of short-term load forecasting depends on the level of the forecasting accuracy, so it’s the key point to study and improve the accuracy in the current time. Domestic and foreign researchers study short-term load forecasting more than long-term forecasting.
On the other hand, with the development of smart distribution grid, the grid will gradually add the smart units, such as variety of distributed power resource, electric vehicle charging station, charging pile and smart electricity-consumption living area, etc. The adoption of these smart units will bring obviously impact to the load model and the load growth model, making the accuracy of conventional forecasting methods can not meet the requirements. To address the situation this paper presents a new method: splitting and modeling. It means split the normal part and the smart part of the grid first, then modeling and forecasting the separated parts. At last, combine each forecasting results parts together, you can get the entire distribution grid power load forecasting results.
This paper adopt the artificial neural network and time series method to study the short-term power forecasting; in the smart distribution grid aspect, the paper focus on the wind power, photovoltaic generation and electric vehicle charging station, reaching the power output of them.
引文
[1]可再生能源中长期发展规划,中华人民共和国国家发展和改革委员会,2007
[2]余贻鑫,新形势下的智能配电网,电网与清洁能源,2009,25(7):1~3
[3]李建芳,盛万兴,孟晓丽,等,智能配电网技术框架研究,能源技术经济,2011,23(3):31~36
[4]李勋,龚庆武,胡元潮,等,智能配电网体系探讨,电力自动化设备,2011,31(8):108~111,126
[5]张铁峰,王江涛,苑津莎,智能配电网研究,电力系统通信,2007,28(181):49~52
[6]徐丙垠,李天友,薛永端,智能配电网与配电自动化,电力系统自动化,2009,33(17):38~41,55
[7]牛东晓,曹树华,卢建昌,电力负荷预测技术及其应用,北京:中国电力出版社,1998,1~229
[8]康重庆,夏清,张伯明,电力系统负荷预测研究综述与发展方向探讨,电力系统自动化,2004,28(17):1~11
[9]王晨枫,电力系统负荷预测的技术研究,硕士学位论文,大庆石油学院,2003
[10]廖旎焕,胡智宏,马莹莹,电力系统短期负荷预测方法综述,电力系统保护与控制,2011,39(1):147~152
[11]秦桂芳,伍世胜,基于神经网络的电力系统短期负荷预测,电器开关,2011,2:37~43
[12]Alex D Papalexopoulos,Sfrangyou Hao.An implementation of a neural network based load forecasting model for the EMS,IEEE Transactions on Power System,1994,9(4):1956~1962
[13]张晓,电力系统短期负荷预测研究,硕士学位论文,四川大学,2001
[14]柯贤波,周啸波,惠华,中长期负荷预测中一种改进的人工神经网络方法,西北电力技术,2004,3:17~22
[15]Yuan-Yih Hsu,PhD Chien-Chuen Yang,MSc.Design of artificial neural networks for short-term load forecasting, PartⅡ:Muti-layer feed forward networks for peak load and valley load forecasting,IEEE Proceedings-C,1991,138(5):414~418
[16]邰能灵,侯志俭,李涛,基于小波分析的电力系统短期负荷预测方法,中国电机工程学报,2003,23(1):45~50
[17]Wei Sun,Yinglian Bai.Short-term load forecasting based on wavelet transform and BP neural network, Mechanic Automation and Control Engineering (MACE), 2011 Second International Conference 2011,USA,656~659
[18]D.C.Park, et al,Electric Load Forecasting Using An Artificial Neural Network,IEEE Trans on Power System,1991,6(2):442~449
[19]Wei Sun,Jiangchang LU,Yujun He et al,Application on Neural Network Model Combining Information Entropy and ANT Colony Clustering the Theory for Short-term Load Forecasting,Proceedings of the Fourth International Conference on Machine Learning and Cybernetics, Guangzhou,2005,4645~4650
[20]闫承山,刘永奇,人工神经网络在华北电网负荷预测中的应用,电网技术,1998,22(7):5~9
[21]刘纯,范高峰,王伟胜,等,风电场输出功率的组合预测模型,电网技术,2009,33(13):74~79
[22]潘迪夫,刘辉,李燕飞,基于时间序列分析和卡尔曼滤波算法的风电场风速预测优化模型,电网技术,2008,32(7):82-86
[23]王金翠,基于实测数据的风电场风速和风功率短期预测研究,硕士学位论文,东北电力大学,2010
[24]蒋金良,林广明,基于ARIMA模型的自动站风速预测,控制理论与应用, 2008,25(2):374～376
[25]刘晓光,风力发电系统风力机输出特性的模拟与控制,硕士学位论文,青岛大学
[26]邵璠,孙育河,梁岚珍,基于时间序列法的风电场风速预测研究,电力环境保护,2008,24(8):1～5
[27]丁明,张立军,吴义纯,基于时间序列分析的风电场风速预测模型,电力自动化设备,2005,25(8):32~34
[28]肖永山,王维庆,霍晓萍,基于神经网络的风电场风速时间序列预测研究,节能技术,2007,25(2):106~108,175
[29]Kengo Taniguchi,Kazuto Yukita.Study on Forecast of Time Series of Wind Velocity for Wind Power Generation by Using Wide Meteorological Data, SolarEnergy,1998,63(1):61~68
[30]刘长青,钟水库,张继波,太阳能电池输出特性的研究,广西物理,2007,28(1):39~41
[31]戴聿雯,光伏阵列输出特性研究与预估分析,硕士学位论文,合肥工业大学,2007
[32]胡义华,陈昊,徐瑞东,光伏电池板在阴影影响下的输出特性,电工技术学报,2011,26(1):123~128,134
[33]禹华军,潘俊民,光伏电池输出特性与最大功率跟踪的仿真分析,计算机仿真,2005,22(6):248~252
[34]A Memhet,A A Mohammed,A new model for I-V characteristic of solar cell generators and its applications,Solar energy materials and solar cells,1995,123~132
[35]李卫民,郭金川,张秀泉,等,有机光伏电池输出特性模拟实验研究,深圳大学学报理工版,2008,25(1):82~86
[36]王增新,苏适,田沛,光伏发电预测技术的应用研究,云南电力技术,2077,39(4):27~30
[37]Chakraborty S,Weiss M D,Simoes M G.Distributed intelligent energy management system for a single-phase high-frequency AC microgrid,IEEE Transactions on Industrial Electronics,2007,54(1):97-109
[38]Atsushi Yona,Tomonobu Senjyu,Toshigisa Funabashi.Application of Recurrent Neural Network to Short-Term-Ahead Generating Power Forecasting for Photovoltaic System,Power Engineering Society General Meeting,2007,1~6
[39]陈昌松,段善旭,殷进军,基于神经网络的光伏阵列发电预测模型的设计,电工技术学报,2009,24(9):153~158
[40]夏德建,电动汽车研究综述,节能与环保,2010.22(7):49~55
[42]雷黎,刘权彬,电动汽车使用对电网负荷曲线的影响初探,电机技术,2000(1):37~39
[43]刘潇潇,廉国海,电动汽车充电设施商业化运营模式研究,湖南电力,2011,31(1):59~62