Automatic Component Calibration and Error Diagnostics for Model-Based Accelerator Control

51025-98-I Automatic Component Calibration and Error Diagnostics for Model-Based Accelerator Control--Vista Control Systems, Inc., 1137 18th Street, , Los Alamos, NM 87544-3012; (505) 662-2484 Dr. Carl Stern, Principal Investigator Dr. Peter Clout, Business Official DOE Grant No. DE-FG03-98ER82708 Amount: $74,999 Accurate models can play a crucial role in commissioning particle accelerators and storage rings used in high energy physics research and in achieving effective operational control of these facilities. To realize and maintain an accurate model, errors in the model must be identified and corrected. Until now, model calibration and error correction has been done manually and is extremely time consuming due primarily to the large search space. The purpose of this project is to automate model calibration and error correction. It will build on existing technology for model calibration, adding a knowledge-based layer embodying expertise in pattern recognition and data interpretation necessary to apply orbit fitting and data analysis methodologies effectively. In Phase I, a prototype knowledge-based system for automating the use of model calibration codes will be developed. This technology will be tested against a set of simulated problems from anactural research environment. Using this study as a basis, a careful analysis of the requirements for automating model calibration and error diagnosis will be made, including the feasibility of implementing self-correcting online models. Commercial Applications and Other Benefits as described by the awardee: Benefits include enhanced control capabilities in the operations of linear accelerators, colliders, and synchrotron rings, as well as an analytic tool to assist in the future commissioning of new experiments. This research is also expected to lead to technologies for adaptive models and more effective model-based control._