| 基于神经网络回归分析组合模型的能源消耗预测研究 |
| |
| 引用本文： | 陈玉金,刘建永,李凌,伍中军. 基于神经网络回归分析组合模型的能源消耗预测研究[J]. 兵工自动化, 2008, 27(11): 59-60 |
| |
| 作者姓名： | 陈玉金 刘建永 李凌 伍中军 |
| |
| 作者单位： | 解放军理工大学工程兵工程学院,江苏 南京,210007;解放军理工大学工程兵工程学院,江苏 南京 210007;防化指挥工程学院指挥1系,北京 102205 |
| |
| 摘 要： | 通过对时间序列数据进行处理，分别建立回归分析预测模型和BP神经网络预测模型，在此基础上建立基于2种预测方法的组合预测模型。采用熵值法确定组合预测模型的权系数。结合某省能源消耗总量数据进行仿真，结果与实际数据的误差较小，和2种单一的预测方法相比，预测结果更接近于实际情况。
|
| 关 键 词： | 神经网络 回归分析 组合预测 能源消耗 |
Forecast and Research on Energy Consumption Based on Analytical and Combined Model of BP Neural Network and Regression |
| |
| CHEN Yu-jin,LIU Jian-yong,LI Ling,WU Zhong-jun. Forecast and Research on Energy Consumption Based on Analytical and Combined Model of BP Neural Network and Regression[J]. Ordnance Industry Automation, 2008, 27(11): 59-60 |
| |
| Authors: | CHEN Yu-jin LIU Jian-yong LI Ling WU Zhong-jun |
| |
| Affiliation: | CHEN Yu-jin, LIU Jian-yong, LI Ling, WU Zhong-jun (1. Engineering College of Engineering Corps, PLA University of Science ＆ Technology, Nanjing 210007, China; 2. No. 1 Department of Command, Command ＆ Engineering Institute of Chemical Defense, Beijing 102205, China) |
| |
| Abstract: | By processing time sequence data,regression analysis model and BP neural network model were individually built.Based on the two models,a combined forecasting model was established.Entropy value method was applied to ascertain the weights parameters.The energy consumption data of one province was simulated.Its results have little error.Compare with the two single methods,the hybrid measure results is better fit the actual process. |
| |
| Keywords: | BP neural network Regression analysis Hybrid forecasting Energy consumption |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
