Optimized ANN-GA Model for Predicting the Compressive Strength of High-Performance Concrete

In new construction and when mixed with conventional concrete, there is a unique material called UHPC. Significant autogenous shrinkage, poor workability for large-scale production, high prices, and an unpredictable peak curing process are some of the obstacles that prevent UHPC from being widely used, despite its remarkable strength. Many large-scale restorations, strengthening projects, long-span bridges, and wet joint connections have made good use of its high compressive strength. Continuous hydration causes autogenous shrinkage to induce early-stage cracking by reducing internal humidity and increasing capillary tension, which in turn causes stress concentration and crack initiation. A new model for HPC modelling that combines ANN and GA is presented in this paper. While PCA chooses important features to feed into deep learning models, the model uses min-max normalisation for preprocessing. Before incorporating GA to find the best process parameters, an ideal ANN design is created utilising a BPNN with BR. When it comes to optimising parameters and forecasting penetration depth, the suggested ANN-GA model outperforms traditional approaches with an accuracy of 98.36%. In order to overcome obstacles in large-scale building projects and enhance HPC applications, this method shows how sophisticated computational approaches can be used.