粉煤灰高性能混凝土的非线性神经网络方法预测

由于废渣高性能混凝土的原料产地多、质量波动大,影响因素复杂,甚至存在诸多因素的交互作用,目前尚难用传统的统计或回归分析方法加以精确描述,本文引入非线性人工神经网络处理技术,利用已有的混凝土配合比（输入）及实测性能指标（输出）等试验数据来训练网络,建立混凝土输入-输出模式非线性映射关系,构造混凝土性能（主要是强度）模型,进一步提升混凝土某些性能的预测精度。

Property Prediction of High Performance Fly Ash Concrete Based on Nonlinear Artificial Neural Networks Method

Meanwhile,because of many raw materials,high quality fluctuation and a great deal of complicated reciprocation influence factors,it is too difficult to describe accurately high performance waste residue concrete with traditional statistics or the return analytical method at present.This text introduces the treatment technology of the non-linear artificial neural network,utilizing mixing ratio of existing concrete（import）and data as the performance index（output）to train the network,setting up concrete to non-linear import——Output mode relation,constructing concrete performance model,promoting prediction precision of concrete performance further.