Tunnel fan embedded foundation damage identification method based on attention measurement convolutional neural network

Abstract

The invention relates to a tunnel fan embedded foundation damage identification method based on an attention measurement convolutional neural network, which belongs to the field of tunnel fan detection. The method comprises the following steps of S1, carrying out a vibration test through a tunnel jet fan foundation health detection system, and collecting excitation and response vibration signals of a tunnel jet fan, S2, performing Fourier transform on the one-dimensional vibration time domain signal to obtain a frequency domain signal, further obtaining a transfer function, and dividing the transfer function into a training set and a test set, S3, building a tunnel jet fan embedded foundation damage identification model based on an MIA-CNN network, training the network by using a training set, continuously adjusting parameters, and performing model optimization, S4, using the test set for tunnel jet fan embedded foundation damage identification, and verifying the performance of the tunnel jet fan embedded foundation damage identification model, and S5, performing tunnel jet fan embedded foundation damage identification by using the verified tunnel jet fan embedded foundation damage identification model.

Description

technical field [0001] The invention belongs to the technical field of tunnel fan detection, and relates to a method for identifying damage to pre-embedded foundations of tunnel fans based on a metric attention convolutional neural network. Background technique [0002] With the rapid development of road traffic, a large number of road tunnels have been built, and road tunnels are generally equipped with mechanical ventilation devices, and more than 95% of them use suspended jet ventilation, so the stability of the suspended jet fan embedded foundation has been highly recognized by people. Pay attention to. [0003] Such as figure 1 As shown, the embedded parts of the jet fan in the highway tunnel include the embedded steel bar A0 set in the concrete structure, the embedded steel plate A1 welded on the embedded steel bar A0, and the mounting bracket A2 welded on the embedded steel plate A1. Fix the jet fan A3 on the mounting bracket A2 with bolts. Since the fan is heavy a...

AI Technical Summary

Problems solved by technology

[0005] However, the existing detection methods all have defects: (1) for the pull-out test, because the fan is large in size and generally has a diameter of more than one meter, it is very difficult to operate the pull-out test under the condition that the fan has been installed, and the observation is also difficult. very difficult

If the fan is removed and the pull-out test is performed, the workload will be large and the test period will be too long

(2) Non-destructive testing is used to mainly detect the reliability of the connection between the embedded steel plate and the mounting bracket. In fact, it is difficult to detect the reliability of the connection between the embedded steel plate and the embedded steel bar, and it cannot detect the embedded steel bar and concrete looseness of the bond

This method can solve the problem of model universality caused by differences in the self-weight, eccentricity, and working conditions of the fan, and reliably detect the health status of the foundation of the embedded parts. Identify or artificially judge the stability of the wind turbine foundation. The judgment result depends on the experience accumulated by the engineer, which is highly subjective and easily leads to misjudgment of the result.

Therefore, it is very difficult to popularize and apply this technology, and it cannot be widely used in the application of stability detection of suspended fan foundations in highway tunnels, and its detection efficiency and intelligence are also low.

Images