BP neural network prediction of concrete compressive strength based on mixing production data
In order to solve the problems of long testing period and lagging of project management in concrete production,a BP neural network model for predicting concrete compressive strength based on real-time monitoring data of concrete mixing production was proposed.Taking the 8 items of material production weighing data and 5 items of production mixing ratio data in concrete mixing production as input variables,200 sets of production monitoring data and test data sample sets were established,which were divided into training set,verification set,and test set according to the ratio of 6∶2∶2.The 8 items of material weighing data and all 13 items of data of C40 ratio concrete mixing and production were used as input variables for 28 d compressive strength prediction of concrete,and the predicted results were compared with the actual test results to verify the effectiveness of the proposed BP neural network prediction model.The results show that the proposed BP neural network concrete strength prediction model can effectively predict the 28 d compressive strength of concrete in real time,and the prediction relative error is better than the estimated value using 7 d compressive strength test data.The BP neural network prediction model with 8 items of material weighing data as input variables has better prediction accuracy,with a mean absolute percentage error of 0.82%and a root mean square error of 0.52 MPa.Furthermore,the adaptability of the BP neural network concrete compressive strength prediction model with 8 items of input variables is verified using production monitoring data of C20 and C30 ratios from different mixing stations,and the prediction mean absolute errors are within 0.5 MPa,and the average absolute percentage errors are less than 2%,consistent with the prediction errors of C40 ratio.The prediction model fully explores the value of real-time monitoring data of concrete mixing station production,realizes the advance of the traditional concrete compressive test results,and has important practical significance for improving the quality level of engineering construction.
concreteprediction modelBP neural networkcompressive strengthproduction monitoring data
NETL
NSTL
万方数据
