新拌混凝土坍落过程的智能预测方法研究  

Research on intelligent prediction method of fresh concrete slump process

作　　者：刘广鹤 王琛 丁然[1] 唐亮 樊健生[1] LIU Guanghe;WANG Chen;DING Ran;TANG Liang;FAN Jiansheng(Department of Civil Engineering,Tsinghua University,Beijing 100084,China;China Construction Engineering Industry Technology Research Institute Co.,Ltd.,Beijing 101300,China)

机构地区：[1]清华大学土木工程系,北京100084 [2]中建工程产业技术研究院有限公司,北京101300

出　　处：《建筑结构》2024年第24期99-106,98,共9页Building Structure

基　　金：国家重点研发计划项目(2022YFC3802000);国家自然科学基金项目(52293433)。

摘　　要：混凝土浇筑质量对复杂混凝土结构和组合结构的力学性能有重要影响,传统的研究方法通常采用足尺浇筑试验,耗时耗力。基于计算流体力学(CFD)方法、机器学习模型XGBoost和视频深度生成模型SimVP,提出了新拌混凝土坍落过程的智能预测方法。首先使用了CFD软件模拟混凝土的坍落度试验,生成坍落过程中的速度云图,为预测模型提供数据集;随后分别构建了典型混凝土配合比(含拌和后时间)-流变参数(屈服应力和塑性粘度)、流变参数(屈服应力和塑性粘度)-坍落过程的智能预测模型。结果表明:智能预测模型能够准确预测混凝土的坍落过程,预测结果在扩展度和流速两项指标上与CFD方法的模拟结果较为一致,且预测速度相比CFD方法提升了4~8倍。The quality of concrete pouring has an important influence on the mechanical properties of complex concrete structures and composite structures.The traditional research methods usually use full-scale pouring test,which is time-consuming and labor-consuming.Based on computational fluid dynamics(CFD)method,machine learning model XGBoost and video depth generation model SimVP,an intelligent prediction method for the slump process of fresh concrete was proposed.Firstly,the slump test of concrete was simulated by CFD software,and the velocity cloud diagram in slump process was generated,which provides the data set for the prediction model.Then,intelligent prediction model of typical concrete mix ratio(including mixing time)-rheological parameters(yield stress and plastic viscosity)and rheological parameters(yield stress and plastic viscosity)-slump process were constructed respectively.The results show that the intelligent prediction model can accurately predict the slump process of concrete,and the predicted results are in good agreement with the CFD method in the two indexes of slump flow and flow velocity,and the predicted speed is 4~8 times higher than that of CFD method.

关 键 词：混凝土浇筑 CFD模拟 人工智能 视频预测 混凝土坍落 

分 类 号：TU528[建筑科学—建筑技术科学]