| 基于多智能算法混凝土强度预测系统设计 |
| |
| 引用本文： | 习华勇,赵林明,习华元. 基于多智能算法混凝土强度预测系统设计[J]. 微计算机信息, 2007, 23(6): 224-225 |
| |
| 作者姓名： | 习华勇 赵林明 习华元 |
| |
| 作者单位： | 1. 056021,河北邯郸,河北工程大学
2. 050408,平山,平山县滹北灌区 |
| |
| 基金项目： | 河北省建设厅建工新产品基金计划项目 |
| |
| 摘 要： | 由于高强混凝土配方数据构成的复杂性,直接利用神经网络模型对其强度进行预测时存在精确度较差的问题。该文应用聚类分析原理对样本数据进行自身特性分析,然后对其自然分组,在此基础上建立多个BP神经网络拟合模型。强度预测时,根据实际数据自适应选择拟和模型,这样在大样本量混凝土强度预测中得到了较好的精度。
|
| 关 键 词： | 混凝土强度预测 优化设计 BP神经网络 聚类分析 |
| 文章编号： | 1008-0570(2007)02-3-0224-02 |
| 修稿时间： | 2006-01-03 |
Design of concrete strength prediction system based on multi-intelligent algorithms |
| |
| XI HUAYONG,ZHAO LINMING,XI HUAYUAN. Design of concrete strength prediction system based on multi-intelligent algorithms[J]. Control & Automation, 2007, 23(6): 224-225 |
| |
| Authors: | XI HUAYONG ZHAO LINMING XI HUAYUAN |
| |
| Abstract: | It has a low accuracy to forecast concrete strength utilizing NN prediction model directly because relation of sample data describing high-strength concrete mixture ratio is complicated. In this paper, the sample data is grouped according to its distribution rule based on clustering analysis firstly, and then, the NN fitting model is developed for every last sample data grouping. In this way, the forecasting accuracy of concrete strength inferences with smaller margins of error to many sample data has been made. |
| |
| Keywords: | prediction of concrete strength optimum design BP neural network clustering analysis |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
