基于模糊系统方法的混凝土强度预测模型  

Prediction model of concrete strength based on fuzzy system method

作　　者：杨松森[1] 徐菁[1,2] 冯启民[3] 

机构地区：[1]青岛理工大学土木工程学院,山东青岛266033 [2]中国海洋大学计算机科学系,山东青岛266071 [3]中国海洋大学工程学院,山东青岛266071

出　　处：《水运工程》2009年第3期18-23,共6页Port & Waterway Engineering

基　　金：国家自然科学基金项目(资助号:60374031)

摘　　要：采用单值模糊器、乘积推理机、中心平均解模糊器,建立了基于模糊系统方法的混凝土无损检测强度评定模型。为了提高模型精度,以回弹值、超声值、碳化深度值和含水率作为模型输入变量,抗压强度值作为模型输出变量,通过一定的数据训练,提取模糊规则,形成完备的模糊规则库。实验结果表明:模糊系统模型预测结果的平均相对误差为7.98%,相对标准差为10.53%。该模型的预测精度明显高于目前常用的回归模型预测精度,基于多因素的混凝土强度评定精度高于基于单因素的混凝土强度评定精度。因此,该方法为混凝土强度检测评定提供了一种有效的途径。Using the singleton fuzzifier,the product inference engine and the center-average defuzzifier,the concrete strength evaluation model of undamaged inspection is obtained based on fuzzy system method.In order to improve the prediction accuracy of the model,the rebound value,ultrasonic value,carbonation depth value and rate of moisture are used as the inputs of the model,and the concrete compressive strength is used as the output of the model.By the sample data training,fuzzy rules can be extracted.All rules constructed a complete base of fuzzy IF-THEN rules.The experiment shows that the average relative error of the predicted results is 7.89%,and the relative standard error is 10.53%.The predicted accuracy of the fuzzy model is evidently higher than that of the regression model.At the same time,the predicted accuracy of concrete strength evaluation based on multifactor is higher than that of concrete strength evaluation based on single factor.So the method provides an efficient way for the concrete strength evaluation.

关 键 词：无损检测 模糊系统方法 混凝土 抗压强度 

分 类 号：TU528.571[建筑科学—建筑技术科学]