MIX RATIO DESIGN AND PERFORMANCE OF PERMEABLE CONCRETE BASED ON A BP NEURAL NETWORK

As an important part of a sponge city and China's rapid urbanisation process, pervious concrete can be widely used in urban light traffic roads and sewage systems. However, at present, there are many factors affecting the performance of pervious concrete, and the flexural strength of ordinary pervious concrete is poor, so its application is limited. A BP neural network model serves as the foundation for the study, which is optimised using the Tianniu Xu search algorithm and the Levenberg Marquardt algorithm in order to address the aforementioned problems. At the same time, the factors influencing the performance of permeable concrete are analysed using a multi-factor and multi-level experimental result visualisation analysis approach. According to the study's findings, the enhanced BP neural network prediction model may reach an optimal convergence state with an error of just 1.663 × 10-6 after just 25 iterations. After the BP-M2VA analysis, the optimal pervious concrete mix ratio of ST1 is 13.95 kg, of ST2 is 7.3 kg, the cement content is 4.30 kg, and the water consumption is 1690 mL. In summary, the analysis method proposed in this study can effectively predict and optimise the mix ratio of perishable concrete, prepare perishable concrete materials with both low cost and high efficiency, and then improve the ecology and environment of the urbanisation process, enhance the safety and comfort of motorists, and promote the sustainable development of urban traffic.