Published Data
These pages provide an access point to data contained in CCFE published journal papers. By selecting a paper, and then a specific figure or table, you can request the related underlying data if it is available for release.
Publication Figures
Publication Date:
2020-02-04
First Author:
V. Gopakumar
Title:
Image Mapping the Temporal Evolution of Edge Characteristics using Neural Networks
Paper Identifier:
Search the list of Figures using keywords (case insensitive):
To drill down to information on Data within a published figure, select a table cell where the figure's number of Data Items > 0
Use the Download buttons to download information on each listed figure, e.g. the image file and how it was created
| Figure Reference | Title | Description | Number of Figure Data Items | Identifier | Download Figure Details | ||
|---|---|---|---|---|---|---|---|
| This image is the first image in the paper, giving an overview on the results of our work. It is referred to as Figure 1 in our paper. | Comparing the Temporal Evolution of Electron Density from SOLPS and the Neural Network Framework | Mapping the temporal evolution of electron density. The set of images on top depicts the evolution from the initial state to the final state as solved by the SOLPS framework. Similarly the set in the bottom, represents the evolution characterised by our novel Fully Convolutional Neural Network approach. | 0 | CF/19/231 | Download | ||
| The figure is referred to as Figure 2 in the paper. | Structural outline of the devised novel Neural Network. | Structure designed to model plasma and neutral evolution using fully convolutional neural networks. | 0 | CF/19/232 | Download | ||
| Referred to as Figure 3 in our paper. | Information Exchange within SOLPS | Information between Plasma and Neutral States exchanged between B2.5 and EIRENE within SOLPS. | 0 | CF/19/233 | Download | ||
| Referred to as figure 4(a) in the paper. | Poloidal Meshgrid Discretisation : SOLPS | Discretisation of the Poloidal Space for a Single Null Configuration as mentioned in the SOLPS manual (referred to in the paper) | 0 | CF/19/234 | Download | ||
| The figure is referred to as figure 4(b) in the paper. | Poloidal Domain Split into Operational Zones : SOLPS | The poloidal region of interest, being split into 4 operational regions. Plot obtained from the SOLPS manual (referred to in the paper). | 0 | CF/19/235 | Download | ||
| Referred to as Figure 5 in the paper. | Numerical Rectangular Grids | Rectangular Grids which form the computational space of SOLPS modelling, adapted to our case that models a JET case. | 0 | CF/19/236 | Download | ||
| Referred to as Figure 6 in the paper. | Network Structure : Internal Configuration | Network Structure exposing the internal configuration of the various layers, indicating the data processing across the network. | 0 | CF/19/237 | Download | ||
| Referred to as Figure 7 in the paper. | PCA of the SOLPS dataset | Principal Component Analysis performed on the SOLPS dataset utilising Singular value Decomposition. The dataset includes rectangular profiles of the plasma and neutral density, temperature and parallel velocity along the tokamak edge. This method was done to showcase the variedness of the dataset which we used to train the network. | 0 | CF/19/238 | Download | ||
| This figure is referred to as figure 8a in the paper. | Electron Density Evolution | Contour Plots of the Electron Density profiles as generated by SOLPS and our Neural Network. The SOLPS solution is plotted as filled contour plots in green and blue, while the Neural Network solution is over-imposed onto the latter in contour plots ranging from yellow to red. | 0 | CF/19/239 | Download | ||
| Referred to as figure 8b in the paper. | Ion Density Evolution | Contour Plots of the Ion Density profiles as generated by SOLPS and our Neural Network. The SOLPS solution is plotted as filled contour plots in green and blue, while the Neural Network solution is over-imposed onto the latter in contour plots ranging from yellow to red. | 0 | CF/19/240 | Download | ||
| Referred to as figure 9a in the paper. | Electron Temperature Evolution | Contour Plots of the Electron Temperature profiles as generated by SOLPS and our Neural Network. The SOLPS solution is plotted as filled contour plots in green and blue, while the Neural Network solution is over-imposed onto the latter in contour plots ranging from yellow to red. | 0 | CF/19/241 | Download | ||
| Referred to as figure 9b in the paper. | Ion Temperature Evolution | Contour Plots of the Ion Temperature profiles as generated by SOLPS and our Neural Network. The SOLPS solution is plotted as filled contour plots in green and blue, while the Neural Network solution is over-imposed onto the latter in contour plots ranging from yellow to red. | 0 | CF/19/242 | Download | ||
| Referred to as figure 10a in the paper. | Evolution of Parallel Velocity of Ions | Contour Plots of the profiles of Ion Parallel Velocities as generated by SOLPS and our Neural Network. The SOLPS solution is plotted as filled contour plots in green and blue, while the Neural Network solution is over-imposed onto the latter in contour plots ranging from yellow to red. | 0 | CF/19/243 | Download | ||
| Referred to as figure 10b in the paper. | Evolution of Parallel Velocity of Neutrals | Contour Plots of the profiles of Neutral Parallel Velocities as generated by SOLPS and our Neural Network. The SOLPS solution is plotted as filled contour plots in green and blue, while the Neural Network solution is over-imposed onto the latter in contour plots ranging from yellow to red. | 0 | CF/19/244 | Download | ||
| Referred to as figure 11 in the paper. | Neutral Density Evolution | Contour Plots of the Neutral Density profiles as generated by SOLPS and our Neural Network. The SOLPS solution is plotted as filled contour plots in green and blue, while the Neural Network solution is over-imposed onto the latter in contour plots ranging from yellow to red. | 0 | CF/19/245 | Download | ||
| Referred to as figure 12 in the paper. | Performance of various FCN achitectures | Scatter plot measuring the impact of depth of the FCN on the performance, with the number of trainable parameters indicated by the marker size. | 0 | CF/19/246 | Download | ||
| Referred to as figure 13 in the paper. | Convolutional Strategies | This figure shows how the aspect ratio of the input data was skewed while deploying the convolutional strategy. | 0 | CF/19/247 | Download | ||
| Referred to as figure 14 in the paper. | Performance of Convolutional Strategies | Bar graph depicting the performance of various convolutional strategies expressed in the Mean Squared Error | 0 | CF/19/248 | Download | ||
| Referred to as figure 15 in the paper. | Internal layout of the optimum FCN model | Graph depicting the information flow and processing as it traverses through our optimum FCN. | 0 | CF/19/249 | Download | ||
Loading... please wait...
