| 郭华,陈勇,马耀光.组合灰色预测模型在入库流量预测中的应用[J].干旱地区农业研究,2012,30(3):96~100 |
| 组合灰色预测模型在入库流量预测中的应用 |
| Application of combined grey forecasting model in the prediction of incoming flow |
| |
| DOI：10.7606/j.issn.1000-7601.2012.03.17 |
| 中文关键词: 灰色GM(1,1)模型 BP神经网络 马尔柯夫链 预测模型 入库流量 冯家山水库 |
| 英文关键词:GM(1,1) BP artificial neural network Markov chain prediction model inflow Fengjiashan reservior |
| 基金项目:国家自然科学基金(50879071) |
|
| 摘要点击次数: 453 |
| 全文下载次数: 333 |
| 中文摘要: |
| 本文将灰色GM(1,1)模型、BP人工神经网络和马尔柯夫链相结合,利用历年入库流量及千河径流量建立组合模型对入库流量进行预测.GM(1,1)模型主要预测趋势,其前半部分与实测值拟合较好,BP神经网络模型后半部有波动部分与实测值拟合较好,二者结合使相对误差最小建立组合模型,同时运用马尔柯夫链预测入库流量的变化范围。预测2001和2002年的入库流量对模型进行检验:GM(1,1)模型预测的相对误差分别为0.359和- 0.017; BP神经网络预测的相对误差分别为0.032和-0.251,组合模型相对误差分别为0.164和0.117,组合预测值在预测区间之内,该组合模型预测结果合理有效,能更精确预测冯家山水库入库流量。 |
| 英文摘要: |
| With the combination of gray GM(1,1) model, BP artificial neural network and Markov chain, the in-coming flow to a reservoir was predicted by using the data of past inflow and the Qian river runoff establish combination model. GM(1,1) model mainly prediet trend, the first half and measured data is better, the BP neural network model second half part have volatility and measured data is better, making relative error smallest to establish combination mod-el, and using markov chain predict incoming flow range change. Forecasting infolw of 2001 and 2002 to inspection flow model:the relative error of GM(1,1) model were 0.359 and 0.017; The relative error of BP neural network were 0.032 and 0.251, the relative error model portfolios were 0.164 and 0.117, combined prediction value within the
range. This combination forecasting model is reasonable and effective, and can be more precise forecasts Fengjiashan reservoir inflow. |
| 查看全文 查看/发表评论 下载PDF阅读器 |
| | |