Initial Performance Tradeoffs in a Rectilinear Muon Cooling Channel with Conventional and Surrogate-Assisted Optimizations

Muon colliders are currently under consideration for building a frontier-energy machine for experimental high energy physics. However, achieving the required luminosity demands cooling muon beams generated from targets by up to five orders of magnitude in 6D emittance. The design of an ionization cooling channel that meets these requirements poses an interesting design challenge which may be tackled with multi-objective optimization. We apply genetic algorithms coupled with full-physics simulations to explore performance tradeoffs in these systems. Additionally, we investigate a hybrid approach incorporating machine learning to accelerate the optimization process. Our preliminary results surpass previously reported performance metrics for initial cooling stages [1], suggesting potential for more compact, efficient cooling channel designs in future muon colliders.

[1] Stratakis, D., & Palmer, R. B. (2015). Physical Review Special Topics - Accelerators and Beams, 18(3), 031003