Online storage ring optimization using dimension-reduction and genetic algorithms
14
TL;DR: In this paper, the authors investigate algorithms that restrict the search space to a small number of linear combinations of parameters ("knobs") which contain most of the effect on the chosen objective (the vertical emittance), thus enabling efficient tuning.
read more
Abstract: Particle storage rings are a rich application domain for online optimization algorithms. The Cornell Electron Storage Ring (CESR) has hundreds of independently powered magnets, making it a high-dimensional test-problem for algorithmic tuning. We investigate algorithms that restrict the search space to a small number of linear combinations of parameters ("knobs") which contain most of the effect on our chosen objective (the vertical emittance), thus enabling efficient tuning. We report experimental tests at CESR that use dimension-reduction techniques to transform an 81-dimensional space to an 8-dimensional one which may be efficiently minimized using one-dimensional parameter scans. We also report an experimental test of a multi-objective genetic algorithm using these knobs that results in emittance improvements comparable to state-of-the-art algorithms, but with increased control over orbit errors.
read more
Chat with Paper
AI Agents for this Paper
Find similar papers on Google Scholar, PubMed and Arxiv
Write a critical review of this paper
Analyze citations of this paper to find unaddressed research gaps
Citations
Machine learning for orders of magnitude speedup in multiobjective optimization of particle accelerator systems
TL;DR: In this paper, the authors introduce an approach based on machine learning to create nonlinear, fast-executing surrogate models that are informed by a sparse sampling of the physics simulation, which enables new ways for high-fidelity particle accelerator simulations to be used, at comparatively little computational cost.
106
•Posted Content
Information geometry for multiparameter models: New perspectives on the origin of simplicity.
TL;DR: In this paper, the authors use information geometry to explore sloppiness and its deep relation to emergent theories and introduce the model manifold of predictions, whose coordinates are the model parameters, whose hyperribbon structure explains why only a few parameter combinations matter for the behavior.
20
Visualizing probabilistic models in Minkowski space with intensive symmetrized Kullback-Leibler embedding
Han Kheng Teoh,Katherine N. Quinn,Katherine N. Quinn,Jaron Kent-Dobias,Colin B. Clement,Qingyang Xu,James P. Sethna +6 more
TL;DR: In this article, the authors show that the predicted probability distributions for any $N$-parameter statistical model taking the form of an exponential family can be explicitly and analytically embedded isometrically in a Minkowski space.
9
Information geometry for multiparameter models: new perspectives on the origin of simplicity
28 Dec 2022
TL;DR: In this article , the authors introduce the model manifold of predictions, whose coordinates are the model parameters, and connect the hierarchy of hyperribbon widths to approximation theory, and to the smoothness of model predictions under changes of the control variables.
References
•Book
Multi-Objective Optimization Using Evolutionary Algorithms
Kalyanmoy Deb,Deb Kalyanmoy +1 more
- 01 Jan 2001
TL;DR: This text provides an excellent introduction to the use of evolutionary algorithms in multi-objective optimization, allowing use as a graduate course text or for self-study.
SPEA2: Improving the strength pareto evolutionary algorithm
Eckart Zitzler,Marco Laumanns,Lothar Thiele +2 more
- 01 Jan 2001
TL;DR: An improved version of SPEA, namely SPEA2, is proposed, which incorporates in contrast to its predecessor a fine-grained fitness assignment strategy, a density estimation technique, and an enhanced archive truncation method.
6K
Performance assessment of multiobjective optimizers: an analysis and review
TL;DR: This study provides a rigorous analysis of the limitations underlying this type of quality assessment in multiobjective evolutionary algorithms and develops a mathematical framework which allows one to classify and discuss existing techniques.
Parameter space compression underlies emergent theories and predictive models.
Benjamin B. Machta,Benjamin B. Machta,Ricky Chachra,Mark K. Transtrum,Mark K. Transtrum,James P. Sethna +5 more
TL;DR: An information-theoretical approach is used to distinguish the important parameters in two archetypical physics models and traces the emergence of an effective theory for long-scale observables to a compression of the parameter space quantified by the eigenvalues of the Fisher Information Matrix.
351
Experimental determination of storage ring optics using orbit response measurements
TL;DR: In this article, a computer code called LOCO (Linear Optics from Closed Orbits) was developed to analyze the NSLS X-ray ring measured response matrix to determine: the gradients in all 56 quadrupole magnets; the calibration of the steering magnets and BPMs; the roll of the quadrupoles, steering magnets, and beam position monitors about the electron beam direction; the longitudinal magnetic centers of the orbit steering magnets; and the horizontal dispersion at the orbit steerable magnet; and transverse mis-alignment of the electron orbit in each
329
Related Papers (5)
J. K. Jones
- 01 Jan 2006
J. K. Jones
- 01 Jan 2006
L. Yang
- 28 Mar 2011
Nicholas Walker,J. Irwin,Mark Woodley +2 more
- 17 May 1993