| 小波网络和BP网络在负荷预测中的比较研究 |
| |
| 引用本文： | 冯再勇,钟康惠,马永旺. 小波网络和BP网络在负荷预测中的比较研究[J]. 江汉大学学报(自然科学版), 2007, 35(2): 51-55 |
| |
| 作者姓名： | 冯再勇 钟康惠 马永旺 |
| |
| 作者单位： | 成都理工大学,信息管理学院,成都,610059;苏州机电高等职业技术学校,江苏,苏州,215031;成都理工大学,地球科学学院,成都,610059;成都理工大学,信息管理学院,成都,610059 |
| |
| 基金项目： | 国家“973”计划资助项目(2002CB412607) |
| |
| 摘 要： | 从收敛速度、网络稳定性、泛化性能及预测结果等方面分析BP网络和小波神经网络的预测过程,探讨了差异的原因.发现:(1)BP网络预测稳定性强,易取舍数值,但预测速度较慢,预测精度不够理想;(2)小波网络预测速度快,精度高,泛化性能好,但稳定性能差,实际运用中还没有合适的数值取舍方法;(3)小波网络总体超过BP网络,但实际预测时技术问题较大.
|
| 关 键 词： | BP网络 小波网络 收敛速度 稳定性 泛化性能 预测精度 |
| 文章编号： | 1673-0143（2007）02-0051-05 |
| 修稿时间： | 2006-09-18 |
Comparison of BP and WNN in Power Load Forecasting |
| |
| FENG Zai-yong,ZHONG Kang-hui,MA Yong-wang. Comparison of BP and WNN in Power Load Forecasting[J]. Journal of Jianghan University(Natural Sciences), 2007, 35(2): 51-55 |
| |
| Authors: | FENG Zai-yong ZHONG Kang-hui MA Yong-wang |
| |
| Affiliation: | 1. College of Information Management, Chengdu University of Technology, Chengdu 610059, China; 2. College of Earth Science, Chengdu University of Technology, Chengdu 610059, China; 3. Suzhou Higher Vocational School of Mechatronic Technology, Suzhou 215031, Jiangsu, China |
| |
| Abstract: | Combining with the practical power load forecasting example,analysing the pre-diction courses of BP and WNN in convergence speed,stability,generalization and prediction ac-curacy.It is found that: ① BP model forecasting is better in stability,selects data easier while operates slower,gets inaccurate results;② WNN model predicts faster,brings out more accurate results,generalizes better while stability is worse,has difficult in selecting data;③On the whole,WNN model performs better than BP model while it has some technical problems. |
| |
| Keywords: | BP model WNN model convergence speed stability generation prediction accuracy |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
