再生骨料透水混凝土关键性能统计及预测分析

对再生骨料透水混凝土(RAPC)4项关键性能指标(抗压强度、劈拉强度、孔隙率及透水系数)进行了统计分析,发现这4项性能指标均基本服从正态分布规律;同时建立了RAPC宏观性能的统计规律与内在联系.在此基础上,基于人工神经网络方法,运用Python软件建立了基于BP神经网络的RAPC性能预测模型,并对上述关键性能指标进行了相互预测分析.结果表明:4项性能指标的模型预测值平均相对误差均在10%以内,预测精度较高,表明RAPC的透水性能与强度性能之间具有内在的反向关联关系,并具备可预测性.

Key Performance Statistics of Recycled Aggregate Pervious Concrete and Prediction Analysis

The statistical analysis of 4key performance indicators(compressive strength, splitting tensile strength, porosity and water permeability) of recycled aggregate pervious concrete(RAPC) revealed that these indicators basically obey the normal distribution pattern; based on this, the statistical characteristics and the intrinsic relation of its macroscopic properties were established. On this basis, and based on the artificial neural network method, the backpropagation(BP) network model was established by means of Python software. The model was used to conduct interactive prediction analysis on the key performance indicators of the concrete; the relative error of the average predicted values is within 10%. This indicates that the application of the method can achieve higher accuracy in performance prediction; it indicates that there is an inherent inverse relationship between the permeability and strength properties of RAPC and it is predictable.