Ponente
Sra.
María Pereira Martínez
(Instituto Galego de Física de Altas Enerxías, Universidade de Santiago de Compostela)
Descripción
An optimization framework is presented for a Parallel-Plate Avalanche Counter (PPAC) with Optical Readout for heavy-ion tracking and imaging. In a previous work, a differentiable optimization framework was developed in which a surrogate model predicted reconstructed positions of impinging charged particles as a function of detector parameters. This approach is extended by introducing a generative surrogate that simulate full detector events as produced by Geant4, while the subsequent position reconstruction is formulated as a differentiable step within the optimization pipeline. The performance of several generative models is compared, and their potential for automated detector design is discussed.
Autores primarios
Sra.
María Pereira Martínez
(Instituto Galego de Física de Altas Enerxías, Universidade de Santiago de Compostela)
Pietro Vischia
(Universidad de Oviedo)
Xabier Cid Vidal
(USC - IGFAE)
