A view of the experimental setup of SST-1’s soft x-ray tomography system. Seven and five cameras are placed around one poloidal cross section of the torus.
Simulated brightness values with mode , modeled on SST-1 equilibrium. “●” data with 2% noise added. “—” fit provided by the reconstructed data by second order linear regularization algorithm.
Reconstruction of simulated emissivity with mode modeled on SST-1 equilibrium. Top row is the simulated emissivity distribution. Second row shows the reconstruction from artificial data without noise, whereas the bottom row is the reconstruction with 2% noise added. Reconstructions are done using first order linear regularization, second order linear regularization, minimum fisher information and maximum entropy algorithms (left to right).
SVD analysis of the time series of simulated emissivity distribution with mode incorporated. Top left figure shows the Singular Values against the index ; whereas top right figure shows the plot of the maximum emissivity as a function of time. Middle row from left to right shows the spatial eigenmodes or topos for . Bottom row shows the corresponding temporal modes or chronos.
Same as Fig. 4 for reconstructed emissivity. Here no noise has been incorporated in the simulated brightness values. Second order linear regularization algorithm has been used for reconstruction.
Same as Fig. 4 for reconstructed emissivity; 2% noise has been incorporated in simulated brightness values.
Comparison of topos number 2 using different algorithms and different noise levels. Top figure is the simulated distribution. Second row is the reconstruction from data without noise. Bottom row shows the reconstruction with 2% noise in brightness values.
Emissivity distribution obtained from seven array SXR tomography system. (a) Simulated emissivity using of size . (b) Reconstructed emissivity without noise and (c) reconstructed emissivity with 2% noise in brightness values. Second order linear regularization algorithm has been used for reconstruction.
Performance test of different algorithms using square pixels. CPU time and reconstruction errors in emissivity have been compared for different methods with 0% and 2% noise incorporated in brightness values.
Performance evaluation by SVD analysis of simulated emissivity with mode and reconstructed emissivity by different reconstruction algorithms. are the relative energies of the first, second and third topos/chronos pairs.
Article metrics loading...
Full text loading...