基于串联灰色神经网络的电力负荷预测方法  

Power Load Forecasting Method Base on Serial Grey Neural Network

作　　者：张大海[1] 毕研秋[1] 毕研霞 毕研梅 牛兆水 洛鲁宁 

机构地区：[1]山东大学电气工程学院,山东济南250061 [2]山东华能德州发电厂,山东德州253024

出　　处：《系统工程理论与实践》2004年第12期128-132,共5页Systems Engineering-Theory & Practice

摘　　要：　为了提高电力负荷预测的精度,分析现有人工神经网络和灰色预测方法各自的优缺点,将二者相结合提出了一种串联灰色神经网络预测方法.新方法利用灰色预测中的累加生成运算对原始数据进行变换,从而得到规律性较强的累加数据,便于神经网络进行建模和训练,同时避免了灰色预测方法存在的理论误差.最后实际算例证明了方法的有效性.方法适用于中长期负荷预测.To improve the accuracy of power load forecasting, we analyze the merits as well as defects of artifical neural network (ANN) and grey prediction method, and we combine the two methods to propose a novel forecasting method named serial grey neural network. The new method utilizes the acuumulation generation operation of grey prediction to transform the origianl data, and produces the accumulated data which possesses better regularity, making it easier to model and train the ANN, and it avoids the theoretical error of grey prediction method. Case study validates the effectiveness of the proposed method, and it is suitable for middle and long term load forecasting.

关 键 词：电力系统 负荷预测 人工神经网络 BP神经网络 灰色系统 

分 类 号：TM715[电气工程—电力系统及自动化]