This file upload contains a GUI (FiberSesorLeakDetectionREV2.mlapp) which serves to connect Picoscope with Matlab to collect signal from our sensor, as well as to perform signal analysis onto the collected signals to identify the leak location on a leaky steel pipeline.
Included, is the PicoScope3000SeriesAApiProgrammersGuide.pdf which explains the functions used to control Picoscope in test_2channels_GUI.m
testanalysiscrtnewfld.m is the signal analysis script. The analysis of the signal starts with performing wavelet de-noising to remove the background/noise profile in the leak signal. The analysis is then continued with pulse conversion where the de-noised signal is converted into a series of square waves using thresholding method. The threshold used here is based on the study of the amplitude of the collected leak signals that indicates the first impact of the acoustic wave onto the sensor. Note that the amplitude is not the maximum amplitude of the signal, instead it is a value very close to zero.
The purpose of this conversion is to make the collected leak signals on both channels/ from both sensors identical, while still maintaining the time lag in between them, so that the cross correlation results can be improved and can more accurately pinpoint the location of the leak cavity on a leaky pipeline.
The cross correlation graph in the GUI is to show the cross correlation results over 9 repetitions of signal collection,in this instance. The variance of the cross correlation results below a certain value means the signal collected are signals from a true leak.
Amanda SL (2019). Acoustic Emission Detection by Fibre Sensor for Pipeline Leak Location Detection (https://www.mathworks.com/matlabcentral/fileexchange/64149-acoustic-emission-detection-by-fibre-sensor-for-pipeline-leak-location-detection), MATLAB Central File Exchange. Retrieved .