The “calculation” axis will focus on modeling for all types of accelerators, required in all stages of machine design and construction. Computation and modeling are the foundation of R&D, and the multi-physical nature of accelerator physics opens up a broad panorama for this cross-disciplinary theme of the network. Numerous codes and software programs are used and developed by the community for various applications: beam physics, electromagnetism and RF, source and plasma physics, vacuum, thermal, cryogenics, etc. The resolution methods (analytical computation, tracking, and particle-in-cell methods) and the required resources (memory, CPU, etc.) also vary greatly.
The computational tools and needs of the community will be the subject of particular attention starting in 2024 and will be addressed through the organization of a workshop in 2024. The objective will then be to strengthen exchanges within the community to define synergies and collaborations between stakeholders involved in different projects but facing similar challenges.
Optimizing control systems and ensuring operational reliability is common to all machines in operation. Much work, based on the use of machine learning methods or the use of advanced algorithms (genetic type), is already underway in laboratories. Although the objectives are diverse (anomaly detection, optimization of beam lifetime, optimization of simulation time, etc.), the mathematical tools are very generic. The field of artificial intelligence is very broad, and scientific exchange within workshops or dedicated working groups will allow a more precise definition of needs in the different fields and will contribute to the development of a national roadmap on the subject.