Muhammad Hanafi, Yusop (2023) Exploring the optimal potential of transient reflection method through mel-frequency ceptrums coefficient and artificial neural network for leak detection and size estimation in water distribution systems. PhD thesis, Universiti Malaysia Pahang Al-Sultan Abdullah (Contributors, Thesis advisor: Mohd Fairusham, Ghazali).
|
Pdf
ir.MUHAMMAD HANAFI BIN YUSOP_PMV 18006.pdf - Accepted Version Download (454kB) | Preview |
Abstract
Water pipeline systems are critical infrastructures that provide potable water to communities. The design and operation of these systems are complex and require careful consideration of various factors, such as system reliability. Regular maintenance and inspection of pipelines and other components are necessary to prevent leaks and ensure that the system operates effectively. The efficient detection and accurate estimation of leaks in water distribution systems are crucial for maintaining the integrity and functionality of the infrastructure. This research aims to unleash the full potential of the transient reflection method through the integration of Mel-Frequency Cepstral Coefficients (MFCC) and Artificial Neural Network (ANN) techniques for leak detection and size estimation in water distribution systems. By leveraging the combined power of signal processing and machine learning, this study aim to advance the state-of-the-art methodologies for leak detection and size estimation, providing more accurate and efficient approaches based on transient reflection method. The objectives of this research are to explores the application of MFCC as a signal processing technique to extract vital information from the transient reflection signals. The transient reflection signals carry valuable insights into the characteristics of the water distribution system and can aid in identifying leaks. Furthermore to investigate and select significant features derived from the transient reflection signals that reflect the nature of leak size. Finally, is to develop and validate an ANN-based model for leak size estimation that harnesses the power of the extracted TRM features. To achieve these objectives, extensive experimentation and analysis will be conducted using transient reflection method obtained from laboratory scale water distribution systems. The data will be collected from various sizes of leaks. The collected dataset will serve as the foundation for training and validating the developed ANN model. Performance evaluation metrics, such as accuracy, precision, recall, and mean squared error, will be utilized to assess the effectiveness and reliability of the leak detection and size estimation technique. The expected outcomes of this research include advancements in leak detection and size estimation techniques in water distribution systems. The integration of MFCC and ANN techniques has the potential to significantly improve the accuracy and efficiency of leak detection, leading to timely identification and mitigation of leaks. The developed estimation model can aid in assessing the severity of leaks, enabling more effective allocation of resources for repair and maintenance activities. Ultimately, the findings of this research will contribute to the enhancement of water distribution system management, promoting water conservation and minimizing the adverse impacts of leaks on infrastructure and the environment. In conclusion, this research endeavors to unleash the full potential of the transient reflection method through the integration of MFCC and ANN techniques for leak detection and size estimation in water distribution systems. By leveraging signal processing and machine learning, this study aims to advance the state-of-the-art methodologies and provide more accurate and efficient approaches to address the challenges associated with leak detection and size estimation. The outcomes of this research have the potential to significantly benefit water management authorities, utilities, and researchers working in the field of water distribution system management and conservation.
| Item Type: | Thesis (PhD) |
|---|---|
| Additional Information: | Thesis (Doctor of Philosophy) -- Universiti Malaysia Pahang – 2023, NO. CD: 13479, SV: Assoc. Prof. Ir. Dr. Mohd Fairusham Bin Ghazali |
| Uncontrolled Keywords: | Mel-Frequency Cepstral Coefficients (MFCC), Artificial Neural Network (ANN) |
| Subjects: | T Technology > T Technology (General) T Technology > TJ Mechanical engineering and machinery |
| Faculty/Division: | Institute of Postgraduate Studies Faculty of Mechanical and Automotive Engineering Technology |
| Depositing User: | Mr. Nik Ahmad Nasyrun Nik Abd Malik |
| Date Deposited: | 06 Jun 2024 04:44 |
| Last Modified: | 06 Jun 2024 04:44 |
| URI: | http://umpir.ump.edu.my/id/eprint/41488 |
| Download Statistic: | View Download Statistics |
Actions (login required)
 |
View Item |