A φ-otdr Vibration Signal Recognition Algorithm Based on stft-cnn-rvfl

Abstract

The present invention relates to the Φ-OTDR vibration signal identification algorithm based on STFT-CNN-RVFL, a method for identifying and classifying the Φ-OTDR vibration signal time-frequency diagram, belonging to the field of image processing and pattern recognition, characterized in that the following steps are adopted: (1) Perform STFT transformation on the Φ-OTDR vibration signal to obtain the time-frequency map; (2) Do grayscale processing on the time-frequency map; (3) Build a CNN network to extract image features; (4) Randomly initialize the connection weights and Threshold; (5) construct RVFL neural network; (6) calculate output weight. The invention takes the time-frequency image of the vibration signal as input, realizes the automatic extraction of vibration signal features through the convolutional neural network, and combines the powerful image recognition and classification functions of the convolutional neural network into the vibration signal recognition. Experimental results show that the vibration signal recognition algorithm designed by the present invention has a better signal recognition effect, and provides an accurate recognition algorithm for the field of vibration signal recognition.

Description

technical field [0001] The invention relates to the field of picture processing and pattern recognition, and mainly relates to a method for recognizing and classifying time-frequency diagrams of Φ-OTDR vibration signals. Background technique [0002] At present, for the classification of Φ-OTDR vibration signals, it is mainly realized by using traditional machine learning algorithms. The general processing flow is to first denoise the signal, decompose it, extract useful features, and finally train the model according to the extracted features to achieve classification. However, Due to the complex and changeable signals, in the process of extracting signal features, low time-frequency accuracy and false component interference may occur, which will affect the accuracy of subsequent signal classification. Although the traditional neural network relying on the gradient descent method has good generalization ability, the convergence speed is too slow during the model training pr...

AI Technical Summary

Problems solved by technology

It uses the gradient descent method to minimize the error between the actual output value and the expected output value during each training process, thereby approaching the objective function, but often falls into a local minimum

Images