| 基于神经网络优化组合的短期负荷预测 |
| |
| 引用本文： | 陈晓娥,康龙云. 基于神经网络优化组合的短期负荷预测[J]. 计算机与数字工程, 2010, 38(3): 161-164 |
| |
| 作者姓名： | 陈晓娥 康龙云 |
| |
| 作者单位： | 1. 陕西工业职业技术学院,咸阳,712000
2. 华南理工大学,广州,510641 |
| |
| 摘 要： | 通过对电力负荷变化规律和影响因素的分析,集结多种单个模型所包含的信息,进行最佳组合,提出了在单一模型预测结果基础之上的基于神经网络的优化组合预测,确定了网络训练样本和隐含层的个数,可使提前一天的预测精度较传统预测模型有较大提高。并当发现某一点预测误差过大,可对该点利用文中提出的误差灰色模型修正预测结果,这样不仅可提高整体预测精度,更重要的是减小最大预测误差值和减少大预测误差发生的次数。仿真结果验证了该预测模型的可行性和有效性。
|
| 关 键 词： | 短期负荷预测 神经网络 灰色模型 预测精度 |
Short-term Load Forecasting Based on Optimized Combination with Neural Network |
| |
| Chen Xiao'e,Kang Longyun. Short-term Load Forecasting Based on Optimized Combination with Neural Network[J]. Computer and Digital Engineering, 2010, 38(3): 161-164 |
| |
| Authors: | Chen Xiao'e Kang Longyun |
| |
| Affiliation: | Shaanxi Polytechnic Institute1;South China University of Technology2 |
| |
| Abstract: | An optimized combination forecast based on neutral network utilizing single model forecast results has been developed.Through the analysis of the electric load change rules and the impact factor,we utilize the information of several kinds of single models and apply the best combination to determine the network training samples and the number of the hidden layer.The forecast accuracy of one day earlier model has been greatly improved,compared with conventional forecast model.At the same time,the Error Grey M... |
| |
| Keywords: | short-term load forecasting neural network grey model forecasting accuracy |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
