Fermilab, Batavia, Illinois, USA
Northwestern University, Evanston, Illinois, USA
The Real-time Edge AI for Distributed Systems (READS) project’s goal is to create a Machine Learning (ML) system for real-time beam loss de-blending within the accelerator enclosure, which houses two accelerators: the Main Injector (MI) and the Recycler (RR). In periods of joint operation, when both machines contain high intensity beam, radiative beam losses from MI and RR overlap on the enclosure¿s beam loss monitoring (BLM) system, making it difficult to attribute those losses to a single machine. Incorrect diagnoses result in unnecessary downtime that incurs both financial and experimental cost. The ML system will automatically disentangle each machine¿s contributions to those measured losses, while not disrupting the existing operations-critical functions of the BLM system. Within this paper, the ML models, used for learning both local and global machine signatures and producing high quality inferences based on raw BLM loss measurements, will only be discussed at a high-level. This paper will focus on the evolution of the architecture, which provided the high-frequency, low-latency collection of synchronized data streams to make real-time inferences.
Performed at Northwestern with support from the Departments of Computer Science and Electrical and Computer Engineering
Fermilab, Batavia, Illinois, USA
The current program at Fermilab involves the construction of a new superconducting linear accelerator (LINAC) to replace the existing warm version. The new LINAC, together with other planned improvements, is in support of proton beam intensities in the Main Injector (MI) that will exceed 2 MW. Measuring the transverse profiles of these high intensity beams in a ring requires non-invasive techniques. The MI uses ionization profile monitors as its only profile system. An alternative technique involves measuring the deflection of a probe beam of electrons with a trajectory perpendicular to the proton beam. This type of device was installed in MI and initial studies of it have been previously presented. This paper will present the status and recent studies of the device utilizing different techniques.
Fermilab, Batavia, Illinois, USA
The Proton Improvement Plan - II (PIP-II) is a new linear accelerator (LINAC) complex being built at Fermilab. It is based on superconducting radiofrequency cavities and will accelerate H− ions to 800 MeV kinetic energy before injection into the existing Booster ring. Measurements of the profile of the beam along the LINAC must be done by non-intercepting methods due to the superconducting cavities. The method chosen is photoionization of a small number of H− by a focused infrared laser, aka laserwire. The number of ionized electrons is measured as a function of laser position within the H− beam. To aid in the design of the collection mechanism, a simulation was written in MATLAB with input from the commercial electromagnetic simulation, CST. This simulation calculates the number and positions of the liberated electrons and tracks them through the magnetic collection and H− beam fields to the collection point. Results from this simulation for various points along the LINAC will be shown.
Fermilab, Batavia, Illinois, USA
One of the on-going issues with the use of microchannel plates (MCP) in the ionization profile monitors (IPM) at Fermilab is the significant decrease in gain over time. There are several possible issues that can cause this. Historically, the assumption has been that this is aging, where the secondary emission yield (SEY) of the pore surface changes after some amount of extracted charge. Recent literature searches have brought to light the possibility that this is an initial ’scrubbing’ effect whereby adsorbed gasses are removed from the MCP pores by the removal of charge from the MCP. This paper discusses the results of studies conducted on the IPMs in the Main Injector at Fermilab.