CSTNet: A Dual-Branch Convolutional Neural Network for Imaging of Reactive Flows Using Chemical Species Tomography

Jiang, Yunfan; Si, Jingjing; Zhang, Rui; Enemali, Godwin; Zhou, Bin; McCann, H.; Liu, Chang

doi:10.1109/tnnls.2022.3157689

Cited by 8 publications

(7 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Therefore, a good fitting result indicates a well-maintained spectral integrity in the down-sampled WMS-2f /1f, and thus high-accuracy projection data, i.e., path In the 4-second experiment, the fuel flow of burner 1 is decreased gradually from 0 to 2 nd second and then increased from 2 nd to 4 th second, while the fuel flow of burner 2 remains stable. The temperature images with 48×48 pixels are reconstructed by CSTNet [30]. Given the imaging rate of 1 kfps, 4,000 frames of temperature images are reconstructed.…”

Section: Experiments Validationmentioning

confidence: 99%

FPGA-Accelerated Distributed Sensing System for Real-Time Industrial Laser Absorption Spectroscopy Tomography at Kilo-Hertz

Xia

Enemali

Zhang

et al. 2024

IEEE Trans. Ind. Inf.

Self Cite

View full text Add to dashboard Cite

Fast and continuous data acquisition (DAQ)with well resolved spectral information is essential for highspeed and high-fidelity measurement of thermophysical parameters of industrial processes using laser absorption spectroscopy tomography (LAST). However, the state-ofthe-art DAQ systems suffer a) inability to collect raw spectral data in real time due to the very high data throughput; b) degradation of spectral integrity when excessive on-chip down-sampling is implemented to reduce data throughput. In this work, we designed a star-networked and reconfigurable DAQ system for real-time LAST imaging at kilo-Hz frame rate. The DAQ system is embedded with a new field programmable gate array (FPGA)-accelerated digital lock-in (DLI) technique, whereby a cascaded integrator-comb (CIC) filter is implemented for down-sampling of the raw signal with well-maintained spectral information. Furthermore, a customized data-encapsulation protocol is developed to enable continuity of real-time data communication between the front-end DAQ hubs and back-end processor. Performance of the developed DAQ system is experimentally validated by flame temperature imaging at 1 kHz, providing the necessary temporal resolution to penetrate turbulent flow and related industrial processes such as reaction propagation.

show abstract

Section: Experiments Validationmentioning

confidence: 99%

FPGA-Accelerated Distributed Sensing System for Real-Time Industrial Laser Absorption Spectroscopy Tomography at Kilo-Hertz

Xia

Enemali

Zhang

et al. 2024

IEEE Trans. Ind. Inf.

Self Cite

View full text Add to dashboard Cite

show abstract

“…The RoS is divided into two parts: RoI that contains the target and the out-of-RoI region that reflects the background. It is worth mentioning that the dimension of RoI that covers the target flame can be determined in advance in both the lab-scale test [20] and industry application [15]. Although the flow field itself is dynamic, it can be located within the RoI.…”

Section: B Size-adaptive Hybrid Meshing Schemementioning

confidence: 99%

Spatially Driven Chemical Species Tomography With Size-Adaptive Hybrid Meshing Scheme

Zhang

Enemali

et al. 2022

IEEE Sensors J.

Self Cite

View full text Add to dashboard Cite

This paper develops a size-adaptive hybrid meshing scheme for Chemical Species Tomography (CST) that is driven by the customized spatial resolution of the sensing region. Traditionally, the entire sensing region in CST is uniformly discretized with the empirically determined density of the meshes. Such a discretization results in a) waste of computational efforts on the less spatially resolved location; and b) much severer rank deficiency. To solve the above-mentioned issues, we introduce, for the first time, a size-adaptive hybrid meshing scheme for CST. Driven by the spatial resolution, dense meshes are deployed in the region of interest (RoI) to detail the target flow field while sparse ones are deployed out of the RoI to fully consider the physically existing laser absorption. The proposed scheme is numerically validated using a CST sensor with 128 laser beams. The visual and quantitative metric comparisons show that the proposed hybrid-size meshing scheme outperforms the traditionally uniform-size meshing scheme, giving 35% lower image error and 38% less significant dislocation at a typical 35 dB signal-to-noise ratio in the RoI. The proposed hybrid-size meshing scheme significantly facilitates the reconstruction of the industrial combustion processes where the combustion zone is bypassed by cooling flows. In these scenarios, the proposed scheme can adapt a finer resolution to detail the combustion zone, while maintaining the integrity of the physical model by less resolved reconstruction of the bypass flows.

show abstract

“…Each sample contains an 8*8 Electrical Capacitance Image (ECI) which represents capacitance measurements and a true local liquid hold-up derived from level meter measurements using ( 8) and (9). ECI could better reflect the nature of the symmetrical geometry of ECT sensors compared with the conventional capacitance measurement vectors method [38,49]. Fig.…”

Section: A Dataset Constructionmentioning

confidence: 99%

Flooding Prognostic in Packed Columns Based on Electrical Capacitance Tomography and Convolution Neural Network

Chen

Liu

Yang

et al. 2022

IEEE Trans. Instrum. Meas.

Self Cite

View full text Add to dashboard Cite

The flooding of packed columns is accompanied by a steep increase in liquid hold-up and pressure drop, resulting in lower mass transfer efficiency and potential damage to equipment. This study aims to investigate, for the first time, the feasibility of Electrical Capacitance Tomography and Convolutional Neural Networks as an intensified alternative to conventional flooding prediction methods. Electrical Capacitance Tomography allows variations in the predominant characteristics of flooding events to be investigated in greater detail than in previous research. Combined with Convolutional Neural Networks, the Electrical Capacitance Tomography sensor enables high accuracy on liquid hold-up calculation and strong robustness against noisecontaminated measurements. In this work, a detailed comparison is made between liquid hold-up results using Convolutional Neural Networks and a more conventional Electrical Capacitance Tomography method based on Maxwell equation. Both methods can accurately calculate the liquid hold-up at low gas flow rates. The liquid hold-up predicted according to Maxwell equation did not match the measured values at high gas flow rates, showing discrepancies of up to 68%. In contrast, Convolutional Neural Networks is much superior to the Maxwell equation method at high gas flow rates, giving only 1% mean of difference than the reference liquid hold-up. Electrical Capacitance Tomography supported by Convolutional Neural Networks shows great fidelity for non-invasive monitoring of local liquid hold-up, allowing for more accurate, localized prediction of loading point and flooding point in packed columns.

show abstract

CSTNet: A Dual-Branch Convolutional Neural Network for Imaging of Reactive Flows Using Chemical Species Tomography

Cited by 8 publications

References 39 publications

FPGA-Accelerated Distributed Sensing System for Real-Time Industrial Laser Absorption Spectroscopy Tomography at Kilo-Hertz

FPGA-Accelerated Distributed Sensing System for Real-Time Industrial Laser Absorption Spectroscopy Tomography at Kilo-Hertz

Spatially Driven Chemical Species Tomography With Size-Adaptive Hybrid Meshing Scheme

Flooding Prognostic in Packed Columns Based on Electrical Capacitance Tomography and Convolution Neural Network

Contact Info

Product

Resources

About