|
|||||
|
|
| 基于一周内负荷方差分析的96点负荷预测模型研究 | |
| 引用本文： | 秦海超,王玮,周晖,刘景星.基于一周内负荷方差分析的96点负荷预测模型研究[J].华北电力技术,2006(12):6-8,15. |
| 作者姓名： | 秦海超 王玮 周晖 刘景星 |
| 作者单位： | 1. 北京朗新信息系统有限公司厦门分公司,福建,厦门,361000 2. 北京交通大学电气工程学院,北京,100044 3. 赤峰电业局,内蒙古,赤峰,024000 |
| 摘 要： | 不同于通常在96点负荷预测时对工作日和休息日所使用的自然区分方法,本文通过引入方差概念对一周各天同时间段电量进行分析,给出了重新划分工作日和休息日的方法。在满足样本相似性的前提下增加了合格样本的数量,为人工神经网络法96点负荷预测提供更为充分和可靠的历史数据。实际预测计算表明,该方法有效地提高了预测精度。 |
| 关 键 词： | 96点负荷预测 人工神经网络 方差 |
| 文章编号： | 1003-9171（2006）12-0006-03 |
| 收稿时间： | 2006-06-26 |
| 修稿时间： | 2006-06-26 |
Study of Short-term Load Forcasting Model of Every 15 Minutes Based on Variance Analysis of Week Load Data |
|
| Qin Hai-chao,Wang Wei,Zhou Hui,Liu Jing-xing.Study of Short-term Load Forcasting Model of Every 15 Minutes Based on Variance Analysis of Week Load Data[J].North China Electric Power,2006(12):6-8,15. | |
| Authors: | Qin Hai-chao Wang Wei Zhou Hui Liu Jing-xing |
| Affiliation: | 1. Xiamen Branch of Beijing Langxin Information System Co. Ltd. ,Xiamen 361000,China; 2. Electric Engineering Institute of Beijing Jiaotong University,Beijing 100044 ,China ;3. Chifeng Power Supply Bureau,Chifeng 024000,China |
| Abstract: | Differing from the usual method dividing a week into work days and rest days in the load forcasting of every 15 Minutes,this paper,using the idea of variance,based on the analysis of the electric quantities of the same slot of each day during one week,puts forward a new method dividing a week into work days and rest days.Presupposing to satisfy the similarity of samples,the total number of samples up to standard has increased.The more full and reliable history data are provided for load forcasting of every 15 minutes with artificial neural network algorithm.The improvement of forcasting accuracy has been proven by practise. |
| Keywords: | load forcasting of every 15 minutes artificial neural network variance |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ | |