25 April 2022 in IEEE Access
This paper investigates applying artificial intelligence (AI) algorithms to attitude control system of satellites to optimally tune the controller using high performance computing. This methodology is applied to the Virtual Telescope for X-ray Observation mission, which is a precise formation of two separate spacecraft observing multiple objects in the space in the X-ray domain. The mission is divided into phases based on the instrumentation and the mission goal. To reach an stable precise formation robust to stochastic slew and slew rate (i.e., Euler angles and angular velocities) in a minimal constrained time T, consumed energy of the attitude control system, denoted as E, and root-mean-square state error of attitude control system, denoted as e, are minimized. Monte-Carlo simulation is used for the sensitivity analysis of optimization and designing a controller. Deep neural networks (DNN), Gaussian processes (GP), and support vector regression (SVR) learn this optimization as a surrogate model, while their hyperparameters are optimized in a novel approach. THETA supercomputer at Argonne Leadership Computing Facility (ALCF) is used for optimizing the hyperparameters of DNN. The surrogate model meets the requirements of the mission, and it shows a better performance over the optimization and Monte-Carlo. The optimal DNN can satisfy the mission requirements e and T while reducing E for 90% compared to the other given methods.
24 June 2022 in Physical Review Accelerators and Beams
Experiments were performed at the Fermilab Accelerator Science and Technology facility to elucidate the effects of long-range wakefields in TESLA-type superconducting rf cavities. In particular, we investigated the higher-order modes (HOMs) generated in the eight, nine-cell cavities of a cryomodule (CM) due to off-axis steering with correctors located upstream of the CM. We have observed correlated submacropulse centroid slews of a few hundred microns and centroid oscillations at in the rf beam-position-monitor data after the CM. The entrance energy into the CM was 25 MeV, and the exit energy was 100 MeV with and in 50-bunch pulse trains. These experimental results were evaluated for machine learning training aspects which will be used to inform the commissioning plan for the Linac Coherent Light Source-II injector CM. With an entrance beam energy of in this case (or for at the European X-ray Free-electron Laser injector), the HOM-effect mitigation would be particularly critical.