scholarly journals A Modified Leakage Localization Method Using Multilayer Perceptron Neural Networks in a Pressurized Gas Pipe

2019 ◽  
Vol 9 (9) ◽  
pp. 1954 ◽  
Author(s):  
Qi Wu ◽  
Chang-Myung Lee
Keyword(s):  
Neural Networks ◽  
Time Delay ◽  
Multilayer Perceptron ◽  
Cross Correlation ◽  
Peak Frequency ◽  
Flexural Mode ◽  
Location Method ◽  
Ae Signal ◽  
Sensor Signals ◽  
Leakage Location

Leak detection and location in a gas distribution network are significant issues. The acoustic emission (AE) technique can be used to locate a pipeline leak. The time delay between two sensor signals can be determined by the cross-correlation function (CCF), which is a measure of the similarity of two signals as a function of the time delay between them. Due to the energy attenuation, dispersion effect and reverberation of the leakage-induced signals in the pipelines, the CCF location method performs poorly. To improve the leakage location accuracy, this paper proposes a modified leakage location method based on the AE signal, and combines the modified generalized cross-correlation location method and the attenuation-based location method using multilayer perceptron neural networks (MLPNN). In addition, the wave speed was estimated more accurately by the peak frequency of the leakage-induced AE signal in combination with the group speed dispersive curve of the fundamental flexural mode. To verify the reliability of the proposed location method, many tests were performed over a range of leak-sensor distances. The location results show that compared to using the CCF location method, the MLPNN locator reduces the average of the relative location errors by 14%, therefore, this proposed method is better than the CCF method for locating a gas pipe leak.

FAST TIME DELAY NEURAL NETWORKS

2005 ◽  
Vol 15 (06) ◽  
pp. 445-455 ◽  
Author(s):  
HAZEM M. EL-BAKRY ◽  
QIANGFU ZHAO
Keyword(s):  
Neural Networks ◽  
Time Delay ◽  
Frequency Domain ◽  
Cross Correlation ◽  
Input Pattern ◽  
Fast Time ◽  
New Approach ◽  
Entire Data ◽  
Speed Up ◽  
Simulation Results

This paper presents a new approach to speed up the operation of time delay neural networks. The entire data are collected together in a long vector and then tested as a one input pattern. The proposed fast time delay neural networks (FTDNNs) use cross correlation in the frequency domain between the tested data and the input weights of neural networks. It is proved mathematically and practically that the number of computation steps required for the presented time delay neural networks is less than that needed by conventional time delay neural networks (CTDNNs). Simulation results using MATLAB confirm the theoretical computations.

scholarly journals A novel method for quantifying periodicity and time delay in dynamic neural networks using unstable subaction potential threshold depolarizations

2020 ◽  
Vol 123 (3) ◽  
pp. 1236-1246
Author(s):  
Julian Sorensen ◽  
Nick J. Spencer
Keyword(s):  
Neural Networks ◽  
Correlation Function ◽  
Time Delay ◽  
Action Potentials ◽  
Cross Correlation ◽  
Time Lag ◽  
Cross Correlation Function ◽  
Electrical Signals ◽  
The Cross

Techniques to identify and correlate the propagation of electrical signals (like action potentials) along neural networks are well described, using multisite recordings. In these cases, the waveform of action potentials is usually relatively stable and discriminating relevant electrical signals straightforward. However, problems can arise when attempting to identify and correlate the propagation of signals when their waveforms are unstable (e.g., fluctuations in amplitude or time course). This makes correlation of the degree of synchronization and time lag between propagating electrical events across two or more recording sites problematic. Here, we present novel techniques for the determination of the periodicity of electrical signals at individual sites. When recording from two independent sites, we present novel analytical techniques for joint determination of periodicity and time delay. The techniques presented exploit properties of the cross-correlation function, rather than utilizing the time lag at which the cross-correlation function is maximized. The approach allows determination of directionality of the spread of excitation along a neural network based on measurements of the time delay between recording sites. This new method is particularly applicable to analysis of signals in other biological systems that have unstable characteristics in waveform that show dynamic variability. NEW & NOTEWORTHY The determination of frequency(s) at which two sources are synchronized, and relative time delay between them, is a fundamental problem for a wide a range of signal-processing applications. In this methodology paper, we present novel procedures for periodicity estimation for single time series and joint periodicity and time delay estimation for two time series. The methods use properties of the cross-correlation function rather than the cross-correlation function explicitly.

scholarly journals Complex analytic signals applied on time delay estimation

2008 ◽  
Vol 6 (1) ◽  
pp. 11-27 ◽  
Author(s):  
Zoran Velickovic ◽  
Vlastimir Pavlovic
Keyword(s):  
Time Delay ◽  
Cross Correlation ◽  
Real Life ◽  
Time Delay Estimation ◽  
Delay Estimation ◽  
Seismic Signals ◽  
Differential Delay ◽  
Sensor Signals ◽  
Passive Localization ◽  
Analytic Signals

In this paper, we present the concept of the time delay estimation based on the transformation of real sensor signals into analytic ones. We analyze the differential time delay values obtained using real seismic signals, simulated complex analytic signals and simulated complex analytic signals with real parts coming from real seismic signals. The simulation results indicate that the application of complex analytic signals leads to reliable computation of the differential time delay. The influence of specific signal parameters on spectral coherence threshold in systems for passive localization and proposed methods for lowering the threshold is analyzed. The computation of all differential time delays with respect to the reference sensor (geophone) is based on the application of Generalized Cross-Correlation (GCC) applied on corresponding analytic signals. The difficulties to select a peak of cross-correlation function that corresponds to true differential time delay when dealing with real signals are significantly reduced if GCC is applied on analytic signals. The efficiency of the proposed technique on differential delay estimation is performed on deterministic and real-life signals.

Time Delay Estimator for Boiler Tube Leakage Localization

2011 ◽  
Author(s):  
Liansuo An ◽  
Peng Wang ◽  
Guoqing Shen ◽  
Jie Shi
Keyword(s):  
Time Delay ◽  
Background Noise ◽  
Cross Correlation ◽  
Signal To Noise Ratio ◽  
Estimation Algorithm ◽  
Time Delay Estimation ◽  
Boiler Tube ◽  
Boiler Furnace ◽  
Direct Energy ◽  
Leakage Location

The inference of strong background noise and reflected by the wall and tube rows surface makes it impossible that justify accurately leakage position employing the characteristic received by multi-channel sensors. It is the ‘bottleneck’ for promoting the accuracy of boiler tube leakage location. The 600MW supercritical boiler model was established, the leakage source propagation process of reflection and attenuation in boiler furnace was simulated by EASE. The approximate signal to noise ratio (SNR) was obtained and the reverberation time was calculated with the squared impulse response integration method on the foundation of simulation. The time delay estimation algorithm PTN, SWITCH derived from PHAT and ML, respectively, are proposed and experiments results revealed the superiority over the classical generalized cross correlation (GCC) method in reverberant and noisy boiler background. Although SWITCH is outperformed by PTN slightly, but the prior knowledge of reverberant energy to direct energy ratio may be hard to obtain in practice and frequencies onset detection is required in PTN method, so the implementation of SWITCH is much easier.

Sign in / Sign up

Export Citation Format

Share Document