基于ANFIS的混凝土空心砌块砌体抗压强度预测  

Estimation of compressive strength of hollow concrete block masonry based on ANFIS model

作　　者：王凤来[1,2] 周强[2] 朱飞[2] 

机构地区：[1]哈尔滨工业大学结构工程灾变与控制教育部重点实验室,黑龙江哈尔滨150090 [2]哈尔滨工业大学土木工程学院,黑龙江哈尔滨150090

出　　处：《武汉大学学报（工学版）》2015年第3期323-326,共4页Engineering Journal of Wuhan University

基　　金：国家科技支撑计划项目(编号:2013BAJ12B03);黑龙江省建设集团有限公司联合科研项目(编号:MH20100436)

摘　　要：为更准确地预测混凝土空心砌块砌体抗压强度,引入自适应神经模糊系统(ANFIS)建立了预估模型.以砌块强度和砂浆强度作为模型输入变量,砌体抗压强度作为模型输出变量值,构建模糊系统模型来描述它们之间的非线性关系.利用统计的国内56组试验数据对模型进行训练和测试,并将试验值与ANFIS模型预测值、现行规范计算模型预测值进行了对比.研究表明:ANFIS模型可以较好地表达抗压强度与其影响因素之间的非线性映射关系,且该模型的预测精度明显高于目前规范计算模型预测精度,可作为混凝土空心砌块砌体抗压强度预测的一种新方法.To predict the compressive strength of concrete hollow block masonry more accurately, the application of the adaptive neural fuzzy system （ANFIS） is explored. The compressive strength of hollow concrete blocks and that of mortar were considered as the input parameters, and the strength of the masonry was the only output parameter. A fuzzy system model was constructed to describe the nonlinear relationship between the input and output variables. Fifty-six available sets of data obtained from the published literature sources were used to train and test the ANFIS model. The results obtained are compared with the values calculated by the Chinese masonry standard. The study results-show that the ANFIS model can bet- ter express nonlinear mapping relations between the compressive strength and its influencing factors, and has higher simulation accuracy than that of current masonry code. Then it is more likely to provide a new method for the strength prediction of hollow concrete block masonry.

关 键 词：混凝土空心砌块 砌体 抗压强度 ANFIS模型 

分 类 号：TU365[建筑科学—结构工程]