基于非参数回归法的掺粉煤灰混凝土抗硫酸腐蚀能力预测  

Prediction Method for Corrosion Resistance of Fly-ash Concrete by Sulfuric Acid Based on Non-parametric Regression

作　　者：谢世华 毛敏 张彩虹 郭会芳 XIE Shi-hua;MAO Min;ZHANG Cai-hong;GUO Hui-fang(Faculty of Architectural Engineering,Kunming University of Science and Technology,Kunming 650500,China)

机构地区：[1]昆明理工大学建筑工程学院,云南昆明650500

出　　处：《软件导刊》2020年第2期23-26,共4页Software Guide

基　　金：国家自然科学基金项目(51078175)。

摘　　要：结合非参数回归技术分析,通过均匀试验设计方法得到的24组稀硫酸腐蚀掺粉煤灰混凝土试件试验数据,研究pH值、水胶比、水泥用量、粉煤灰用量、坍落度5个因素对混凝土腐蚀量的影响以及相应预测方法。结合试验数据,首先采用Pavlick建议的幂函数模型进行参数回归,获得各组试验的腐蚀系数k与时间指数n,然后在此基础上,采用ACE非参数回归技术分析腐蚀系数k与pH值、水胶比、水泥用量、粉煤灰用量、坍落度5个因素的关系,并建立腐蚀预测模型。结果表明:①幂函数模型能很好地反映腐蚀随时间的增长规律;②按浸泡过程中硫酸钙析出时间点不同,可将k、n值分布特征大致划分为3类,采用ACE回归分别建立不同类预测模型,能获得较好的预测结果。The prediction method for corrosion resistance of fly-ash concrete(FAC)by sulfuric acid is investigated by non-parametric regression to study the effects of five influence factors,i.e.pH value,water-binder ratio,cement dosage,fly-ash dosage and slump.The experimental datum was obtained from 24 groups of uniform designed soaking tests of FAC specimens in dilute sulfuric acid.The power function model proposed by Pavlik was first applied to each group and obtained the corrosion coefficient k and time index n.Then,the corrosion prediction mode is established by analyzing the relationship between corrosion coefficient k and the five influence factors mentioned above with the non-parametric regression technique of ACE.Results show that:①The power functions fit the experimental datum well;②The distribution characteristics of k and n can be roughly divided into three categories according to the precipitation time of calcium sulfate during immersion.A better prediction can be obtained by using ACE non-parametric regression for different corrosion types.

关 键 词：硫酸腐蚀 掺粉煤灰混凝土 均匀试验 非参数回归 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]