Fast singular value decomposition combined maximum entropy method for plasma tomography

Abstract

The maximum entropy method (MEM) is a widely used reconstruction algorithm in plasma physics. Drawbacks of the conventional MEM are its heavy time-consuming process and possible generation of noisy reconstruction results. In this article, a modified maximum entropy algorithm is described which speeds up the calculation and shows better noise handling capability. Similar to the rapid minimum Fisher information method, the modified maximum entropy algorithm uses simple matrix operations instead of treating a fully nonlinear problem. The preprocess for rapid tomographic calculation is based on the vector operations and the singular value decomposition (SVD). The initial guess of the sought-for emissivity is calculated by SVD and this helped reconstruction about ten times faster than the conventional MEM. Therefore, the developed fast MEM can be used for intershot tomographic analyses of fusion plasmas.