The Accelerator Operations and Physics (AOP) Group of the Accelerator Systems Division (ASD), Advanced Photon Source (APS) is hiring a postdoctoral researcher for a multi-year appointment. The postdoc researcher will work to develop and deploy advanced AI/ML methods for accelerator tuning and control, as well as beam and equipment diagnostics, with applications to the APS injector and the newly-upgraded APS storage ring. Experimental work on the existing APS injector accelerator complex and the new APS storage ring will be done to test the methods and to assess their usefulness for accelerator operation. The work is supported by an on-going DOE-funded R&D project in which we work collaboratively with colleagues from SLAC and Brookhaven National Laboratories.

Controls of modern, complex accelerators at large-scale scientific user facilities have become increasingly challenging as new machines push technology limits in the quest for higher performance. Using beam-based diagnostics to identify machine errors that affect the performance and tuning available control parameters will be critical in realizing the design performance of future machines, while applying advanced control methods to accelerator optimization and control during operation will be necessary in maintaining high performance. Machine learning techniques have found use in accelerator tuning and control in recent years.

In this R&D project, we will apply existing methods to important tuning and control problems on injectors and storage rings, and explore and develop more efficient and more robust methods to address the challenges. Studies will also be conducted on using ML methods to analyze operational history data in order to extract information to inform accelerator maintenance decisions by detecting and predict component failures. In these studies, we will develop instrumentation based on low-cost, commodity hardware, such as low-resolution infrared imaging inside power supplies and vibration sensors on water pumps.

Required Knowledge, Skills, and Experience:

• PhD completed in accelerator physics, physical sciences, mathematics, computer science, electrical engineering, or a related field with zero to five years of employment experience.

• Strong programming skills.

• Proficiency in the Python programming language.​

• Working knowledge of UNIX or Linux.

Preferred Knowledge, Skills, and Experience:

• Experience with machine learning and accelerator operation.

• Experience working with complex algorithms.

• Experience using computing clusters for simulation and data analysis.

• Experience with accelerator modeling and simulation codes, such as Elegant.​