Health Monitoring of Carbon Fiber Reinforced Building Materials Based on Phase unwrapping Algorithm

During the construction process, building structures are affected by environmental factors such as earthquakes. Engineers cannot monitor real-time data related to the status of building materials, such as whether the structure is complete, whether nodes are effective, and specific displacement situations. Therefore, a hybrid model to monitor carbon fiber reinforced building materials based on the least squares phase unwrapping and deep convolutional neural networks is developed to provide timely warning of damage factors. When constructing the model, deep convolutional neural network is used to effectively extract the features of carbon fiber reinforced building materials, and improve the analysis ability of the building materials monitoring model. The experimental results showed that the hybrid algorithm had higher computational accuracy and smaller errors compared with discrete cosine transform algorithm, fast Fourier transform algorithm, and quality map guided method. When the signal-to-noise ratio of these four algorithms was 6%, the average errors were 0.3147, 0.8462, 0.7836, and 0.8147, respectively. When conducting denoising tests on noisy images, the hybrid algorithm showed good denoising results, with clear images. In addition, empirical analysis was conducted on the constructed hybrid model for health status monitoring. When the iteration was 120, the loss curve gradually smoothed and the loss rate remained between 0.2 and 0.3. When testing the fracture of a single suspension rod, the difference in cable force detected was 115kN, with a variation amplitude of 12.15%. When the hybrid model was tested in training set and test set, the accuracy of building material monitoring was 98.2% and 95.5%, respectively. The above results indicate that the carbon fiber reinforced building material health monitoring hybrid model can timely detect the damage status of building materials and transmit and process data. This study will better monitor the status of key points in future building structures and provide support for emergency response decisions in structures.