Manufactured Sand Concrete Mix Ratio Design And Eco-efficiency Evaluation Of LCA-LCC Based On Machine Learning

The application of manufactured sand concrete(MSC)is of great significance for the construction industry to achieve sustainable development goals.As one of the important indicators of MSC performance,the compressive strength is still in its infancy to carry out intelligent prediction with the help of machine learning methods.In addition,the mix design comprehensively integrated the sustainability requirements of MSC is still insufficient.In this thesis,the MSC mix ratio data set was established by means of literature research,and then the MSC strength prediction and sustainable evaluation were studied based on the machine learning methods.The research contents and results mainly include the following aspects:(1)Based on back-propagation neural network(BPNN),support vector regression(SVR)and extreme learning machine(ELM)algorithms,a MSC strength prediction model was established and verified.It was found that the MSC strength prediction model established based on the SVR algorithm had the best prediction effect.(2)Based on multiple output regression(MOR),the MSC mix ratio prediction model was established and verified.The fitting determination coefficient R2 values of cement and water demand were both greater than 0.9,which met the accuracy requirements.Then,the initial MSC mix ratio data was generated based on the MOR mix ratio prediction model and the SVR intensity prediction model.On this basis,the volume standardization method was used to screen the mix ratio data,and the MSC mix ratio data sets of different intensity levels were obtained,which were used for the subsequent life cycle.(3)Based on the life cycle assessment(LCA)and the life cycle cost(LCC)methods,the optimal MSC mix ratio data for environmental impact and cost were given,the environmental impact and cost changes of MSC with different strength grades of MSC with 1m3/MPa as the functional unit were analyzed.Environmental impact and cost changes.The results show that the raw material stage has the highest environmental impact and cost,and cement production has the highest environmental impact and cost in this stage,while natural coarse aggregate has the highest environmental impact and cost in the transportation stage.At the same time,17 environmental sub-categories of environmental impact were studied and analyzed,and it was found that the impact of climate change on human health accounted for the largest proportion.Finally,it is found that with the continuous improvement of the strength level of MSC,the utilization rate of each material in the concrete gradually increase,and the environmental impact and cost of the unit strength functional unit gradually decreased.(4)The calculation results based on the LCA and LCC were introduced into the eco-efficiency index,and the optimal eco-efficiency mix ratio data points were given.Then,the maximum information correlation method(MIC)was used to analyze and rank the correlations between the 28d compressive strength and eco-efficiency values and the design parameters of the MSC mix ratio.Finally,the SVR prediction model of LCA,LCC,and ecological efficiency(ECO-E)was established,and the Sobol index method was used to analyze the sensitivity of each parameter.The results show that LCA,LCC and ECO-E are the most sensitive to cement,followed by water-cement ratio and aggregate;ECO-E has similar sensitivity to cement and aggregate content.