Automated Data Processing and Analysis of Rain Acoustic Sensor Data
This study performed an automated data processing of rain acoustic sensor data and analyzed the results when compared to the standard tipping bucket rain gauge data. It improved the tedious acoustic data processing that was manually done using the Audacity software. MATLAB programming software was u...
Saved in:
Main Author: | |
---|---|
Format: | text |
Published: |
Archīum Ateneo
2021
|
Subjects: | |
Online Access: | https://archium.ateneo.edu/theses-dissertations/470 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Ateneo De Manila University |
id |
ph-ateneo-arc.theses-dissertations-1596 |
---|---|
record_format |
eprints |
spelling |
ph-ateneo-arc.theses-dissertations-15962021-10-06T05:19:40Z Automated Data Processing and Analysis of Rain Acoustic Sensor Data Aquino, Danilyn Joy This study performed an automated data processing of rain acoustic sensor data and analyzed the results when compared to the standard tipping bucket rain gauge data. It improved the tedious acoustic data processing that was manually done using the Audacity software. MATLAB programming software was used to automatically pull, pre-process and analyze the available sound recordings of rain that were gathered from September – November 2019. Fast Fourier Transform (FFT) algorithm was applied to extract the significant acoustic features. Laravel, a PHP web framework, was used to display the extracted features and the accumulated rain in the website in real time. Microsoft Excel was used for running statistical tests and for modelling the relationship with the tipping bucket data. FFT results from the MATLAB code were compared to the standard frequencies of archived audio files and resulted to an accuracy of 99.9%. Frequency results from Audacity and MATLAB were analyzed. For a dataset of 685 one-minute audio files, the calculated Mean Absolute Percentage Error (MAPE) was only 3.71%. Pearson’s correlation coefficients of 0.74 to 0.99 showed a strong positive relationship between frequency and rain rate. In terms of range, it was analyzed that the frequencies for each classification (Light, Moderate, Heavy, Intense and Torrential) were distinct with each other at a confidence interval of 99%. The results of the analysis showed that the developed Rain Acoustic Sensor (RAS) can complement the currently used tipping bucket rain gauge and can serve as a foundation for the possibility of deploying rain acoustic sensors in high density. 2021-01-01T08:00:00Z text https://archium.ateneo.edu/theses-dissertations/470 Theses and Dissertations (All) Archīum Ateneo n/a |
institution |
Ateneo De Manila University |
building |
Ateneo De Manila University Library |
continent |
Asia |
country |
Philippines Philippines |
content_provider |
Ateneo De Manila University Library |
collection |
archium.Ateneo Institutional Repository |
topic |
n/a |
spellingShingle |
n/a Aquino, Danilyn Joy Automated Data Processing and Analysis of Rain Acoustic Sensor Data |
description |
This study performed an automated data processing of rain acoustic sensor data and analyzed the results when compared to the standard tipping bucket rain gauge data. It improved the tedious acoustic data processing that was manually done using the Audacity software. MATLAB programming software was used to automatically pull, pre-process and analyze the available sound recordings of rain that were gathered from September – November 2019. Fast Fourier Transform (FFT) algorithm was applied to extract the significant acoustic features. Laravel, a PHP web framework, was used to display the extracted features and the accumulated rain in the website in real time. Microsoft Excel was used for running statistical tests and for modelling the relationship with the tipping bucket data. FFT results from the MATLAB code were compared to the standard frequencies of archived audio files and resulted to an accuracy of 99.9%. Frequency results from Audacity and MATLAB were analyzed. For a dataset of 685 one-minute audio files, the calculated Mean Absolute Percentage Error (MAPE) was only 3.71%. Pearson’s correlation coefficients of 0.74 to 0.99 showed a strong positive relationship between frequency and rain rate. In terms of range, it was analyzed that the frequencies for each classification (Light, Moderate, Heavy, Intense and Torrential) were distinct with each other at a confidence interval of 99%. The results of the analysis showed that the developed Rain Acoustic Sensor (RAS) can complement the currently used tipping bucket rain gauge and can serve as a foundation for the possibility of deploying rain acoustic sensors in high density. |
format |
text |
author |
Aquino, Danilyn Joy |
author_facet |
Aquino, Danilyn Joy |
author_sort |
Aquino, Danilyn Joy |
title |
Automated Data Processing and Analysis of Rain Acoustic Sensor Data |
title_short |
Automated Data Processing and Analysis of Rain Acoustic Sensor Data |
title_full |
Automated Data Processing and Analysis of Rain Acoustic Sensor Data |
title_fullStr |
Automated Data Processing and Analysis of Rain Acoustic Sensor Data |
title_full_unstemmed |
Automated Data Processing and Analysis of Rain Acoustic Sensor Data |
title_sort |
automated data processing and analysis of rain acoustic sensor data |
publisher |
Archīum Ateneo |
publishDate |
2021 |
url |
https://archium.ateneo.edu/theses-dissertations/470 |
_version_ |
1715215759159853056 |