Combination of MFCC Feature Extraction and Support Vector Machine for Pipeline Leak Detection

DOI

10.2991/978-94-6463-772-4_45How to use a DOI?

Keywords

Pipeline; Leak; MFCC; Machine Learning; SVM

Abstract

Pipes have an important role in industries that support fluid or gas distribution. Therefore, it is necessary to periodically monitor the condition of the pipe, especially if what is flowed can be flammable or dangerous if it comes out of the pipe. Leaks in the pipe need to be avoided. In the era of the Industrial Revolution 4.0, condition monitoring by utilizing sensors as data collectors that are processed and machine learning as a condition classifier is being developed. The sound produced by the pipe will be extracted using the Mel-Frequency Cepstral Coefficients (MFCC) feature which will be the input for machine learning. MFCC features are considered to have high sensitivity which is expected to result in high model performance. This research will compare the performance of the model used for classification in 2 (two) different environmental conditions, namely laboratory scale and workshop scale. The SVM-based machine learning model can predict the condition of leaks and no leaks as evidenced by the percentage of F1-score performance that reaches 95% at the laboratory scale and 90.95% at the workshop scale. The model has a high percentage of accuracy which is 89.74% at the laboratory scale and 92.17% at the workshop scale. The general objective of this research is to test the performance of the combination of MFCC feature extraction with the Support Vector Machine model in detecting pipe leaks in environments with different noise levels.

Copyright

© 2025 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Feby Anggraeini Rosyifah
AU  - Anindita Adikaputri Vinaya
AU  - Andhika Eko Prasetyo
PY  - 2025
DA  - 2025/07/01
TI  - Combination of MFCC Feature Extraction and Support Vector Machine for Pipeline Leak Detection
BT  - Proceedings of the 10th International Conference on Science and Technology (ICST 2024)
PB  - Atlantis Press
SP  - 505
EP  - 518
SN  - 2352-5401
UR  - https://doi.org/10.2991/978-94-6463-772-4_45
DO  - 10.2991/978-94-6463-772-4_45
ID  - Rosyifah2025
ER  -