Electrical Capacitance Tomography Data Research Articles

On the Comparison of Capacitance-Based Tomography Data Normalization Methods for Multilayer Perceptron Recognition of Gas-Oil Flow Patterns

Normalization is important for Electrical Capacitance Tomography (ECT) data due to the very small capacitance values obtained either from the physical or simulated ECT system. Thus far, there are two commonly used normalization methods for ECT, but their suitability has not been investigated. This paper presents the work on comparing the performances of two Multilayer Perceptron (MLP) neural networks; one trained based on ECT data normalized using the conventional equation and the other normalized using the improved equation, to recognize gas-oil flow patterns. The correct pattern recognition percentages for both MLPs were calculated and compared. The results showed that the MLP trained with the conventional ECT normalization equation out-performed the ones trained with the improved normalization data for the task of gas-oil pattern recognition.

Flow Regime Classification Using Artificial Neural Network Trained on Electrical Capacitance Tomography Sensor Data

The main goal of the presented work is to analyse the performance of the Multi-Layer Perceptron (MLP) neural network for flow regime classification based on sets of simulated Electrical Capacitance Tomography (ECT) data. Normalised ECT data have been used to separately train several MLPs employing various commonly used back-propagation learning algorithms, namely the Levenberg-Marquardt (LM), Quasi-Newton (QN) and Resilient-Backpropagation (RP), to classify the gas-oil flow regimes. The performances of the MLPs have been analysed based on their correct classification percentage (CCP). The results demonstrate the feasibility of using MLP, and the superiority of LM algorithm for flow regime classification based on ECT data.

Pneumatic Transport of Granular Materials in an Inclined Conveying Pipe: Comparison of Computational Fluid Dynamics−Discrete Element Method (CFD−DEM), Electrical Capacitance Tomography (ECT), and Particle Image Velocimetry (PIV) Results

This paper presents three flow patternsdispersed flow, reverse flow, and half-ring flowin the post-bend region of a 45° inclined pneumatic conveying pipe. Solid concentration and velocity distribution were measured using electrical capacitance tomography (ECT), particle image velocimetry (PIV), and a high-speed camera. The axial velocity of particles and solid transverse motion in the pipe cross section were obtained using a modified cross-correlation method from the ECT data for the three flow patterns. The axial velocity profile over the pipe cross section was compared with data obtained from the PIV system. In particular, emphasis has been placed on the characterization of the reverse-flow pattern, where negative values of solid velocities were observed. By analyzing the images captured with the high-speed camera, three distinct regions in the solid phase of the reverse-flow regime may be discerned. A dense region with high solid concentration formed next to the bottom wall of the inclined conveying pipe, while a dilute region with a relatively low solid concentration formed near the center of the pipe. Reverse flow was observed to occur predominantly in the dense region and a transition region between the two. Finally, the dynamic forces acting on single particles were analyzed for the three flow patterns. This analysis showed that the phenomena of reverse-flow and half-ring flow structure formation may be attributed to the effects of electrostatics. This finding has also been validated with results of numerical simulations performed in the present study.

Advanced Statistical Analysis as a Novel Tool to Pneumatic Conveying Monitoring and Control Strategy Development

AbstractBehaviour of powder flow in pneumatic conveying has been investigated for many years, though it still remains a challenging task both practically and theoretically, especially when considering monitoring and control issues. Better understanding of the gas‐solids flow structures can be beneficial for the design and operation of pneumatic transport installations. This paper covers a novel approach for providing the quantitative description in terms of parameter values useful for monitoring and control of this process with the use of Electrical Capacitance Tomography (ECT). The use of Bayesian statistics for analysis of ECT data allows the direct estimation of control parameters. This paper presents how this characteristic parameters estimation can be accomplished without the need for reconstruction and image post processing, which was a classical endeavour whenever tomography was applied. It is achieved using a ‘high‐level' statistical Bayesian modelling combined with a Markov chain Monte Carlo (MCMC) sampling algorithm. Advanced statistics is applied to data analysis for measurements coming from the part of phenomena present in the horizontal section of pneumatic conveyor during slug formation.

Nonlinear image reconstruction for electrical capacitance tomography using experimental data

Electrical capacitance tomography (ECT) attempts to image the permittivity distribution of an object by measuring the electrical capacitances between sets of electrodes placed around its periphery. Image reconstruction in ECT is a nonlinear and ill-posed inverse problem. Although reconstruction techniques based on a linear approximation are fast, they are not adequate for all cases. In this paper, we study the nonlinearity of the inverse permittivity problem of ECT. A regularized Gauss–Newton scheme has been implemented for nonlinear image reconstruction. The forward problem has been solved at each iteration using the finite element method and the Jacobian matrix is recalculated using an efficient adjoint field method. Regularization techniques are required to overcome the ill-posedness: smooth generalized Tikhonov regularization for the smoothly varying case, and total variation (TV) regularization when there is a sharp transition of the permittivity have been used. The reconstruction results for experimental ECT data demonstrate the advantage of TV regularization for jump changes, and show improvement of the image quality by using nonlinear reconstruction methods.

Inversion of electrical capacitance tomography data by simulated annealing: Application to real two-phase gas–oil flow imaging

In this work we apply a highly optimized simulated annealing (SA) inversion method to the reconstruction of permittivity images from real two-phase gas–oil flow electrical capacitance tomography (ECT) data. We test the SA inversion method using several flow regimes generated by varying gas and oil flow rates in a test loop facility. The SA-based permittivity inversions have some advantages over other reconstruction approaches based on linear least-squares inversion: they can find good solutions starting with poor initial models, can easily implement complex a priori information, and do not introduce smoothing effects in the final permittivity distribution model. A major disadvantage comes from the fact that SA is computationally very intensive and leads to relatively slow reconstructions when calculation of the forward problem is not very fast. In this work we employ a linearized and numerically improved forward model based on the use of a sensitivity matrix. We find this novel approach to be faster and more accurate than traditional linear methods.

Pneumatic transport of granular materials through a [formula omitted] bend

In the present study, a pneumatic conveying system incorporating a 90 ∘ bend is investigated. This study employs the use of three non-invasive instruments to measure solids concentration and velocity distribution determination in the pneumatic conveying system. They are namely the electrical capacitance tomography (ECT), particle image velocimetry and phase doppler particle analyzer. Pressure transducers were also used to monitor the pressure drop characteristics along the post-bend vertical pipe region. Two different classes of granular materials, polypropylene beads ( 2600 μ m , Geldart class D) and glass beads ( 500 μ m , Geldart class B), were used to investigate the differences in the flow characteristics for granular particles of various Geldart classes. The experimental results show a constant frequency pulsating flow for polypropylene beads in the dense-phase flow regime. This is illustrated by the visualization, ECT and pressure drop data. For dilute-phase flow regime, both polypropylene and glass beads show a continuous annulus flow structure. Numerical simulation using the Euler–Euler method was also conducted using computational fluid dynamics and the fluid and particle flow characteristics were compared with the experimental data obtained in the present study.

Simulation and experimental validation of a freely bubbling bed of FCC catalyst

A validation and mesh refinement study has been performed for the simulation of a freely bubbling bed of fluid catalytic cracking (FCC) catalyst operating at superficial gas velocities in the range of 0.05 to 0.20 m/s, using the two-fluid computational fluid dynamics (CFD) code Multiphase Flow with Interphase eXchanges (MFIX). The simulation results have been compared to experimental data collected using an electrical capacitance tomography (ECT) imaging system as well as data from the literature. A comparison has been made based on the bed expansion, and bubble diameters and rise velocities. A novel method has been used to extract bubble properties from ECT data and simulation results. It was found that a modified gas–solids drag law, corresponding to an effective particle agglomerate diameter in the range of 135 to 170 μm for FCC particles of actual 75 μm mean diameter, was required to adequately predict the fluidization behavior observed experimentally. These findings support the argument that cohesive interparticle forces lead to agglomeration of FCC catalyst powder and significantly affect the fluidization quality. The model is the first presented in the open literature to successfully apply CFD to obtain realistic predictions of FCC bubbling bed behavior.

Determination of multi-component flow process parameters based on electricalcapacitance tomography data using artificial neural networks

Artificial neural networks (ANNs) have been used to investigate their capabilitiesat estimating key parameters for the characterization of flow processes, based onelectrical capacitance-sensed tomographic (ECT) data. The estimations of theparameters are made directly, without recourse to tomographic images. Theparameters of interest include component height and interface orientation oftwo-component flows, and component fractions of two-component andthree-component flows. Separate multi-layer perceptron networks were trainedwith patterns consisting of pairs of simulated ECT data and the correspondingcomponent heights, interface orientations and component fractions. Thenetworks were then tested with patterns consisting of unlearned simulatedECT data of various flows and with real ECT data of gas–water flows.The neural systems provided estimations having mean absolute errors ofless than 1% for oil and water heights and fractions and less than 10°for interface orientations. When tested with real plant ECT data, themean absolute errors were less than 4% for water height, less than 15°for gas–water interface orientation and less than 3% for water fraction,respectively. The results demonstrate the feasibility of the application of ANNsfor flow process parameter estimations based upon tomography data.

A new image reconstruction method for tomographic investigation of fluidized beds

AbstractIt is important to investigate the behavior of circulating fluidized beds (CFB) for efficient coal combustion. Electrical capacitance tomography (ECT) was applied to the measurement of gas/solids distribution in square CFBs. An iterative method was used to achieve improved image reconstruction over the commonly used linear back‐projection algorithm. The solids concentration profile, temporal variation of average solids concentration, and frequency spectra, which were obtained from the ECT data, are presented.

Electrical Capacitance Tomography Measurements on the Pneumatic Conveying of Solids

We have evaluated the usefulness of electrical capacitance tomography (ECT) as a tool for monitoring pneumatic conveying in horizontal ducts. Power spectra of solids concentration fluctuations obtained from single-plane ECT data were used to identify the various flow regimes, and these were confirmed through visual observation. From single-plane ECT data, the instantaneous and time-averaged distributions of particle concentration over the cross section of the conveying pipe have been determined in various flow regimes. Propagation velocities of patterns were evaluated from cross correlation of twin-plane ECT data. The solids mass flow rate, determined independently by load cell measurements, was found to be roughly proportional to the product of the pattern velocity, the particle density, and the average solids holdup in the pipe, and the proportionality factor depended on the material being transported through the pipe. In our experiments involving flows past a 90° smooth bend, ECT was able to detect significant temporal and spatial nonuniformity in particle concentration in the postbend region.

Hardware implementation of RAM-based neural networks for tomographic data processing

The hardware discussed in the paper employs field programmable logic devices to interface with memory components and, unlike previous implementations, utilises dynamic RAM without compromising performance. The network offers the opportunity to estimate process parameters without recourse to image reconstruction. Tests reveal speedups of 17, 22 and 6 for image reconstruction, void fraction estimation and flow regime classification, respectively, from electrical capacitance tomography data.

Neural Networks for Prediction and Control of Chaotic Fluidized Bed Hydrodynamics

A neural-network-based model that has learnt the chaotic hydrodynamics of a fluidized bed reactor is presented. The network is trained on measured electrical capacitance tomography data. A training algorithm is used that does not only minimize the short-term prediction error but also the information needed to synchronize the model with the real system. This forces the model to focus more on learning the longer term dynamics of the system, expressed in the average multi-step-ahead prediction error and dynamic invariants such as correlation entropy and dimension. The availability of the model is an important step towards control of chaos in gas-solid fluidized beds.