Artificial intelligence techniques for tuning linear induction accelerators

Darrel L Lager,H R Brand,F W Chambers,F E Coffield,William J Maurer,W C Turner

San Francisco, CA, USA（1991）

被引用6|浏览5

摘要

An expert system has been developed that acts as an intelligent assistant for tuning particle beam accelerators. This system is called MAESTRO-Model and Expert System Tuning Resource for Operators. MAESTRO maintains a knowledge base of the accelerator containing not only the interconnections of the beamline components, but also their physical attributes such as measured magnetic tilts, offsets, and field profiles. MAESTRO incorporates particle trajectory and beam envelope models which are coupled to the knowledge base permitting large number of real-time orbit and envelope calculations in the control-room environment. This capability has been used (1) to implement a tuning algorithm for minimizing transverse beam motion, (2) to produce a beam waist with arbitrary radius at the entrance to a brightness diagnostic, and (3) to measure beam energy along the accelerator by fitting orbits to focusing and steering sweeps.<>

查看译文

关键词

collective accelerators,expert systems,particle beam diagnostics,physics computing,maestro,accelerator tuning,artificial intelligence,beam energy,beam envelope,beam waist,beamline components,brightness diagnostic,expert system,field profiles,knowledge base,linear induction accelerators,magnetic tilts,offsets,particle trajectory,real-time orbit,transverse beam motion,algorithms,tuning

上传PDF