| In recent years, China’s demand for construction of resources increases sharply, but the excessive consumption of resources and the generation of construction waste has caused great harm on the economy and environment. The research and application of recycled concrete, which can effectively solve the problem of difficult to deal with construction waste, realize resource recycling, has good effect on socio-economic and environmental effects, has great realistic significance to the sustainable development of economy and society.As the human brain neural network theoretical digital models, artificial neural network is an information processing system which is modeled on the structure and function of brain neural networks established to allow complex logic operations and non-linear relationship mapping to become a reality. Therefore, the research of this article employs the principle of artificial neural network to forecast the compressive strength of recycled concrete.The research work of this paper is in under a grant from the National Natural Science Foundation of China (51378270),Using MATLAB R2012b Neural Network Toolbox programming, respectively, the forecasting models are established by using BP neural network and RBF neural network. After a lot of trial and coMParison of simulation results to determine the optimal network structure and related parameters of RBF network model and BP network model.Establish two kinds of recycled concrete strength prediction model based on neural networks,and respectively, in accordance with the recycled concrete’s mixture ratio (includingcement dosage, the dosage of recycled aggregates, types of recycled aggregates, recycled aggregate grade, ratio of recycled aggregate replacement, water-cement ratio, admixture dosage), predict its7d,28d compressive strength. The results showed that both the established neural network prediction models can accurately predict the rapid regeneration of the compressive strength of concrete, can meet the needs of the actual project.Overall, it is better to using BP network prediction model for the recycled concrete compressive strength prediction. Based on BP network prediction model is established, and discuss the quality of the recycled aggregate, the replace rate, the dosage of cement, and water-cement ratio on the influence of recycled coarse and fine aggregate concrete strength, and found the conclusions are highly consistent with the actual situation. Test proved that artificial neural network prediction and simulation of the recycled concrete strength which can be applied to practical engineering.In reality, Currently, the preparation of recycled concrete requires a lot of repetitive ratio experiments, time-consuming, labor-intensive, the use of artificial neural networks can effectively reduce the number of tests to avoid waste of resources, increase economic efficiency. |
