Damage identification and evaluation model of blast-resistant structure based on machine learning

Explosion-proof structure plays a vital role in protecting the safety of people and property. However, it is a complex and challenging task to identify and evaluate the damage caused by explosion to these structures. Traditional detection methods are time-consuming, laborious and ineffective in detecting hidden damage. In this paper, a damage identification and evaluation model of blast-resistant structure based on machine learning is proposed. The dynamic response data of the structure under the action of explosion is automatically analyzed by using the convolutional neural network (CNN) in deep learning, so as to quickly and accurately identify the location, degree and type of damage. CNN is selected because of its powerful feature extraction ability and efficient recognition ability of complex patterns. It can automatically learn and extract high-level features from original data, which simplifies the complicated process of manually designing features in traditional methods. The effectiveness of the proposed model is verified by the actual explosion experiment. The experimental results show that the CNN model maintains a high level of accuracy in identifying the three types of damage under different explosion intensities, especially in the medium-intensity explosion. In addition, CNN model shows remarkable advantages in damage identification and assessment tasks, and its accuracy reaches 93.87%, far exceeding the methods based on vibration modal analysis and machine learning. This study not only improves the detection efficiency and reduces the risk of human misjudgment, but also provides a scientific basis for the safety management of anti-explosion structures and contributes a new technical means to improve the level of public safety.