Rubber particles and straw powder were used to prepare recycled rubber straw concrete, and the freeze-thaw test was conducted on the recycled rubber straw concrete using the quick-freezing method. The frost resistance of the recycled rubber straw concrete was evaluated by determining the relative dynamic modulus of elasticity, the rate of mass loss, and the flexural strength of the recycled rubber straw concrete in the process of freezing and thawing. SEM was used to observe the microstructure of the recycled rubber straw concrete after the freezing and thawing process. SEM observed the microstructure of recycled rubber straw concrete after freezing and thawing. The effect and mechanism of rubber admixture and straw admixture on the frost resistance of concrete were investigated by microanalysis. Based on the experimental data, the particle swarm algorithm and genetic algorithm were used to optimize the BP neural network to establish the prediction model of recycled rubber straw powder, and the results show that the PSO-BP neural network prediction model established in this paper has good accuracy and stability. It has a good prediction effect on the flexural strength and the number of freeze-thaw cycles of recycled rubber straw concrete under different mixing ratios.
Keywords: BP neural network model; genetic algorithm (GA); mechanical prediction; particle swarm optimization (PSO) algorithm; rubber concrete; straw powder.