Analysis of Accelerator Tuning Algorithms for Integration with an Automatic Planning and Control System

145

Analysis of Accelerator Tuning Algorithms for Integration with an Automatic Planning and Control System--Vista Control Systems, Inc., 134-B Eastgate Drive, Los Alamos, NM 87544-3304; (505) 662-2484
Mr. Michael Kroupa, Principal Investigator
Mr. Peter N. Klout, Business Official
DOE Grant No. DE-FG03-97ER82506
Amount: $74,999

Particle accelerators have moved from being strictly a research tool of nuclear physicists to playing an ever increasing role in industry. Oftentimes the initial startup and reconfiguration of particle beams constitutes a significant portion of available research or production time. Given the large number of possible applications and the complexity of accelerator operations, new facilities are often forced to invest large amounts of time in deriving their own startup and operational procedures. Using an existing intelligent control architecture along with a new control paradigm, this project will survey the human knowledge and expertise used to generate particle beams for use across a variety of accelerator types and applications. This expertise will be used to generate a syntax for entering this knowledge into a generalized database for use by an automated planning and control system. This new paradigm, known as teleo-reactive control, uses an ordered sequence of condition-action pairs to show adaptive, non-deterministic control behavior.
During phase I of this project, accelerator experts will be interviewed about tuning goals and techniques, and the teleo-reactive control paradigm will be used to construct a generalized database of beam-tuning algorithms. This database will then provide the input for developing an automatic planning and control system having the specific goal of tuning to a specified beam while minimizing the total power in tuning and accelerating elements. During phase II of this project, we will complete the beam-tuning database will be completed, and the automated planning and control system will be generalized for use in the accelerator community and industry.

Commercial Applications and Other Benefits as described by the awardee: This project will develop a beam-tuning database for the particle accelerator community. Since many industrial processes share control characteristics with accelerator-tuning tasks, an automated planning and control system could also provide an advanced information tool for industry.