Detection of small gas leaks based on neural networks and D-S evidential theory using ultrasonics

Wang, Tao; Yu, Pei Lin; Huiheng, Xiao

doi:10.1784/insi.2014.56.4.189

Cited by 3 publications

(4 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…According to the literature [27], it is observed that the leakage signal spectrum is mainly concentrated in the ultrasonic range from 30 kHz to 100 kHz, while it is mainly concentrated around 40 kHz under the leakage aperture and pressure conditions set in this paper. Therefore, the acquisition card is set to a sampling rate of 500 kHz, and the quality of the signal is ensured by high sampling rate, as well as the ability to collect a large amount of data in a short period of time.…”

Section: Data Acquisitionmentioning

confidence: 77%

Dy-GNet: a lightweight and efficient 1DCNN-based network for leakage aperture identification

Zhang,

Li,

Liu

et al. 2024

Meas. Sci. Technol.

View full text Add to dashboard Cite

The deep neural network's gas leakage aperture recognition method has high accuracy, but its low computational efficiency due to its complex structure greatly limit its application in resource-limited industrial environments and real-time processing. In this paper, we proposed an efficient recognition method based on 1D convolutional neural network. First, wavelet scattering coefficients with time-frequency information are obtained using the wavelet scattering transform. Second, the dynamic convolution is used to deepen the feature extraction and PConv to speed up the inference time, and the efficient Dy-G module is constructed. Finally, the Dy-G module is stacked to construct Dy-GNet to achieve the classification task. The performance of the model is verified in a noisy environment, and the results show that the model FLOPs are 17.24 M and 83.94% accuracy is achieved at SNR=0 dB, which guarantees the accuracy while speeding up the inference speed with high efficiency.

show abstract

Section: Data Acquisitionmentioning

confidence: 77%

Dy-GNet: a lightweight and efficient 1DCNN-based network for leakage aperture identification

Zhang,

Li,

Liu

et al. 2024

Meas. Sci. Technol.

View full text Add to dashboard Cite

show abstract

“…The frequency band of the ejection noise is very wide, i.e., approximately several hundred kilohertz [ 21 ]. After an in-depth experimental analysis, Ma Dayou et al have provided an empirical formula for the peak frequency of the ejection noise, f p , as [ 22 ]:

…”

Section: Principles Of Ultrasonic Leak Detectionmentioning

confidence: 99%

Gas Leak Location Detection Based on Data Fusion with Time Difference of Arrival and Energy Decay Using an Ultrasonic Sensor Array

Wang

Hong

2018

Sensors

View full text Add to dashboard Cite

Ultrasonic gas leak location technology is based on the detection of ultrasonic waves generated by the ejection of pressured gas from leak holes in sealed containers or pipes. To obtain more accurate leak location information and determine the locations of leak holes in three-dimensional space, this paper proposes an ultrasonic leak location approach based on multi-algorithm data fusion. With the help of a planar ultrasonic sensor array, the eigenvectors of two individual algorithms, i.e., the arrival distance difference, as determined from the time difference of arrival (TDOA) location algorithm, and the ratio of arrival distances from the energy decay (ED) location algorithm, are extracted and fused to calculate the three-dimensional coordinates of leak holes. The fusion is based on an extended Kalman filter, in which the results of the individual algorithms are seen as observation values. The final system state matrix is composed of distances between the measured leak hole and the sensors. Our experiments show that, under the condition in which the pressure in the measured container is 100 kPa, and the leak hole–sensor distance is 800 mm, the maximum error of the calculated results based on the data fusion location algorithm is less than 20 mm, and the combined accuracy is better than those of the individual location algorithms.

show abstract

“…Considering that the fixed frequency of the ultrasonic transducer can be chosen for reasons of convenience, the center frequency of the detection system was set to 40 kHz. Therefore, the acquisition of a 40 kHz frequency ultrasonic signal was used to detect leakage [17].…”

Section: The Principle Of Ultrasonic Gas Leak Detectionmentioning

confidence: 99%

“…In actual tests, the measurement errors are Δ = ± Δ = d p 0.01 mm, 0.05 dB, resulting in ΔL, and the sound pressure localization error can similarly be obtained. The sound pressure localization error, according to equation (17), can be expressed as:…”

Section: Rough Localization Using Ultrasonic Sound Intensitymentioning

confidence: 99%

Gas leak localization and detection method based on a multi-point ultrasonic sensor array with TDOA algorithm

Wang¹,

Wang²,

Yu³

et al. 2015

Meas. Sci. Technol.

Self Cite

View full text Add to dashboard Cite

To resolve the measured target position to determine and locate leak problems with current gas leak detection and localization systems based on ultrasonic technology, this paper presents an improved multi-array ultrasonic gas leak TDOA (time difference of arrival) localization and detection method. This method involves arranging ultrasonic transducers at equal intervals in a high-sensitivity detector array, using small differences in ultrasonic sound intensity to determine the scope of the leak and generate a rough localization, and then using an array TDOA localization algorithm to determine the precise leak location. This method is then implemented in an ultrasonic leak detection and localization system. Experimental results showed that the TDOA localization method, using auxiliary sound intensity factors to avoid dependence on a single sound intensity to determine the leak size and location, achieved a localization error of less than 2 mm. The validity and correctness of this approach were thus verified.

show abstract

Detection of small gas leaks based on neural networks and D-S evidential theory using ultrasonics

Cited by 3 publications

References 0 publications

Dy-GNet: a lightweight and efficient 1DCNN-based network for leakage aperture identification

Dy-GNet: a lightweight and efficient 1DCNN-based network for leakage aperture identification

Gas Leak Location Detection Based on Data Fusion with Time Difference of Arrival and Energy Decay Using an Ultrasonic Sensor Array

Gas leak localization and detection method based on a multi-point ultrasonic sensor array with TDOA algorithm

Contact Info

Product

Resources

About