000 03398ntm a2200289 a 4500
001 vtls000099807
003 KUKTEM
005 20251117113337.0
008 170503t2016 my a f a m 000 0 eng d
020 _aTHE0005231(Local)
039 9 _a201905141239
_bhanafiah
_c201710111541
_daishah
_y201705031024
_zfateeha
040 _aUMP
090 _aFKM .F38 2016 r Thesis
100 0 _aNurul Fatiehah Adnan
245 1 0 _aLeak detection in gas pipeline based on time-frequency analysis /
_cNurul Fatiehah Adnan
260 _aKuantan, Pahang :
_bUMP,
_c2016
300 _axvii, 114 p. :
_bill. (some col.) ;
_c30 cm. +
_e1 CD ROM
500 _aFaculty of Mechanical Engineering
502 _aThesis (Master of Engineering in Mechanical Engineering) -- Universiti Malaysia Pahang – 2016
504 _aBibliography : p. 98-104
520 3 _aPipelines are very important part of any engineering infrastructure. Gas leak is a prominent problem associated with pipelines and it is inevitable in many systems. Prevention of leak is important since gas leak will lead to deficiency and also effects the environment. The safety is one of the main reasons why leak detection is popular in research and development sector. Subsequently, many researchers studied various available methods for leak detection in gas pipeline. One of them is acoustic method. Leak in gas pipelines can be detected and located precisely using certain sensor. However, high fast alarm rate is the main problem in detecting leak. Non-stationary signal is produced by gas flowing in the pipeline, making the analysis of signal becomes more difficult. The scope of this research is detecting and locating leak in straight pipeline and L-bend. The experiments were conducted by injecting the sinusoidal wave into the pipeline and the reflection of the signal is collected by using a microphone. The data is acquired by DASYlab and analysed using Matlab software. The important element in this study is the signal echoes produced by the disturbance of gas flow with leak (defect) or Lbend (feature). Analysing the signal echoes is very closely related to signal processing. In conjunction to that, Hilbert transform (HT) and Ensemble Empirical Mode Decomposition (EEMD) or simply called as Hilbert Huang Transform (HHT) is applied to produce robust results. EEMD is adopted to decompose the signal into IMFs and each IMF is analysed using HT to get instantaneous phase and instantaneous frequency. Besides, the study has shown that HT provides accurate leak location with application of instantaneous phase and instantaneous frequency. Hilbert spectrum (HS) has successfully displayed the energy-time-frequency distribution in proper presentation. The percentage error for all experiment is below 5 percent. In conclusion, HHT is a promising signal processing method that can be used to analyse the echoes acquired from experiments, producing accurate leak location with this method, it is possible to identify leaks and also features in a pipeline network with acceptable errors for both leak and L-bend.
610 2 0 _aFaculty of Mechanical Engineering
_xDissertations
650 0 _aUniversities and Colleges
_xDissertations
650 0 _aTheses
856 4 0 _uhttp://ecollib.ump.edu.my/25855/
_zLibrary access only
999 _aVIRTUA40
_c7090
_d7096
999 _aVTLSSORT0080*0200*0400*0900*1000*2450*2600*3000*5000*5020*5040*5200*6100*6500*6501*8560*9992