Optical fiber sensing system invasion identification and classification method and classifier based on filter bank

Abstract

The invention discloses an optical fiber sensing system invasion identification and classification method and a classifier based on a filter bank. The method comprises the steps that every two sampling values input at present are added to obtain a data vector with the length being N; a convolution window with the length being N is multiplied with a corresponding element of the data vector, and an intermediate vector is obtained; Q sub-channel filter coefficient vectors are multiplied by the intermediate vector, the products are accumulated, and Q filtering results are output; a power feature vector is input to a neural network to be trained and tested, and finally the type of the invasion action is judged. According to the classifier, signals to be filtered are detected through an end point, the starting point of an event is judged, the signals obtained during a period of time starting from the starting point are sent to an AD converter to be sampled, a sample sequence is obtained and enters a DSP device in a parallel data input form, and by means of the interior algorithm processing of the DSP device, the type of the invasion action is finally output. The optical fiber sensing system invasion identification and classification method and classifier have the advantages of being efficient, rapid and high in accuracy, and the false-alarm phenomenon is avoided.

Description

technical field [0001] The invention relates to the technical field of digital signal processing, in particular to a method for identifying and classifying violations of an optical fiber sensing system based on a filter bank and a classifier. Background technique [0002] With the development of optical fiber and optical fiber communication technology [1] maturation of the fiber optic sensor [2] developed rapidly. Perimeter Security System Based on Optical Fiber Sensing Technology [3] It is also gradually being valued in the field of security. with conventional infrared [4] ,electric fence [5] and laser [6] Compared with other security systems, the optical fiber sensing system is very sensitive to pressure and vibration, and has the functions of sensing and signal transmission. It can monitor in real time various disturbances that are directly touched or indirectly transmitted to the optical fiber. Micro vibrations are converted into electrical signals and transmitted...

AI Technical Summary

Problems solved by technology

Currently, the most commonly used action recognition method is based on wavelet decomposition [9] However, the time-consuming and computational load of this method will increase exponentially with the increase in the number of decomposition layers, and the recognition error rate is relatively high, it is not easy to achieve flexible configuration, so it is difficult to meet the requirements of the security system

Images