Postdoc - Digital Twin and Surrogate Modelling of Laser-Electron Acceleration and -Seeded Free-Electron Lasers
HZDR - Helmholtz-Zentrum Dresden-Rossendorf
Dresden, Germany
Your tasks
- Design and train data-driven surrogate models for laser-plasma acceleration (LPA) simulations;
- Devise models to translate between distributions resulting from simulation and experiments;
- Find LPA parameters to control efficiency and resulting electron bunch shape;
- Match simulation results to virtual diagnostics of experimental LPA-FEL setup to understan LPA setup;
- Integrate and unify simulation workflows and data-handling with ML workflows;
- Close communication with experimental colleagues to understand the setup and communicate promising parameters to explore;
- Publication of results in scientific journals and dissemination at conferences.
Your profile
- Completed PhD in the field of Physics (Computational, Particle Physics, Optics, or related), Computer Science with a strong background in machine learning, statistics;
- Experience in adaptive problem-solving and critical analysis of approaches and results (scientific method);
- Proficient in numerical modeling and computational workflows;
- Experience with machine learning (incl. deep learning), statistics and global optimization methods;
- Skilled in data post-processing, analysis, visualization, presentation and (computer-) experiment design;
- Good programming skills in languages such as Python, Julia or C++ (at least two languages);
- Knowledge of modern deep learning frameworks (e.g. pyTorch, Tensor-flow, Jax, ...);
- Experience working in Linux shell/cluster environment and on shared resources;
- Proficient with collaborative software development tools (e.g. version control: Git);
- Independent, investigative working style;
- Actively seeking input and openly sharing results and knowledge in a team.
Apply Now
Don't forget to mention EuroScienceJobs when applying.