引用本文摘 ? 要：為準確預測高強混凝土（HSC）的強度變化規律，調研收集了225組HSC配合比及其28 d強度數據，對統計變量進行Spearman相關性分析，并基于enter法建立了不同膠凝體系HSC的28 d強度與配合比之間的多元非線性回歸模型。分析了不同模型預測值與實際值的散點圖、殘差圖數據分布情況，并結合擬合優度指標和相對誤差指標對模型進行了評價。結果表明：硅灰摻量、水膠比、膠凝材料總量和減水劑摻量對HSC的28 d強度影響效果顯著，影響程度依次遞減，P值均小于0.01；3種膠凝體系HSC強度回歸模型的回歸效果和預測精度均能滿足預測要求?；诒疚哪Ｐ?，可通過原材料配合比分別對C-F-Si-G四元膠凝體系、C-F-G三元膠凝體系和C-F二元膠凝體系HSC的28 d抗壓強度進行預測。 Abstract: The data base of 225 sets of high-strength concrete(HSC) mix ratio and compressive strength curing for 28 d were established for predicting the strength of HSC accurately. Spearman correlation analysis was performed on statistical variables, and 3 multiple nonlinear regression models of different cementitious systems were established based on the enter method for strength prediction of HSC. The data distribution of scatter plots and residual plots of the predicted values and actual values of different prediction models were analyzed, and the goodness of fit index and the relative error index were also analyzed for model evaluation comprehensively. The results show that the dosage of silica fume, water-binder ratio, binder amount and water reducing agent dosage have a very significant effect on the compressive strength of HSC, the degree of importance decreases successively, and the P value is all less than 0.01. The regression results and prediction accuracy of all the prediction models can meet the prediction requirements. Based on this model, the 28 d compressive strength of HSC of the C-F-Si-G quaternary cementitious system, C-F-G ternary cementitious system and C-F dual cementitious system can be predicted through the different models based on material mixing ratio.
甘彬霖,馮旭海,畢經龍,等.基于多元非線性回歸模型的高強混凝土強度預測研究[J].混凝土與水泥制品,2022(2):1-7. GAN B L,FENG X H,BI J L,et al.Research on the Multiple Nonlinear Regression Prediction Model for Compressive Strength of High-strength Concrete[J].China Concrete and Cement Products,2022(2):1-7.