Concrete Structure Identification and Damage Detection Based on Genetic Algorithm Combined with Cluster Analysis

With the advancement transportation industry, concrete structures are widely utilized in construction due to their benefits including cost-effectiveness and easy construction. To improve the accuracy of concrete structure identification and damage detection, this study uses acoustic emission technology to obtain various waveform parameter features in the structure. It uses Back Propagation Neural Network (BPNN) to improve Genetic Algorithm (GA), while combining K-means++ clustering analysis method for mixed damage identification. The outcomes demonstrated that the model’s accuracy in identifying the location of damage was as high as 98.46%, and the accuracy of identifying the degree of damage was 97.23%. In terms of AUC, the model achieved 0.986 with a misclassification rate of only 1.54%. In summary, the research on concrete structure identification and damage detection based on GA combined with clustering analysis significantly improves the accuracy and reliability of concrete structure damage detection, providing a new technical means for health monitoring of concrete structures in engineering practice.