Fig 9 - available via license: Creative Commons Attribution 4.0 International
Content may be subject to copyright.
The systematic and random components of the true simulated quadrupolar field errors, computed as mean μ and rms σ of the errors distribution, compared to their values in the predicted results on validation data
Source publication
Magnetic field errors and misalignments cause optics perturbations, which can lead to machine safety issues and performance degradation. The correlation between magnetic errors and deviations of the measured optics functions from design can be used in order to build supervised learning models able to predict magnetic errors directly from a selectio...
Context in source publication
Context 1
... as expected larger compared to the triplet magnets, since the reconstruction of the error sources located in the arcs is affected by degeneracy stronger compared to the triplet magnets installed around the IPs. Inspecting the reconstruction of particular components of the simulated errors (systematic and random errors as described in Sect. 2), Fig. 9 demonstrates a very good agreement between the systematic components of the simulated and predicted triplet errors, whereas the random part appears to be the biggest contribution to the overall prediction error of the trained estimator. Combining individual triplet magnets also leads to a better agreement between rms values of ...
Similar publications
Biomimicry involves taking inspiration from existing designs in nature to generate new and efficient systems. The feathers of birds which form a characteristic herringbone riblet shape are known to effectively reduce drag. This paper aims to optimise the individual constituent structure of a herringbone riblet pattern using a combination of computa...
Polyp segmentation has accomplished massive triumph over the years in the field of supervised learning. However, obtaining a vast number of labeled datasets is commonly challenging in the medical domain. To solve this problem, we employ semi-supervised methods and suitably take advantage of unlabeled data to improve the performance of polyp image s...
In this paper, we propose a flexible machine learning framework to predict customer lifetime value (CLV) in the Business-to-Business (B2B) Software-as-a-Service (SaaS) setting. The substantive and modeling challenges that surface in this context relate to more nuanced customer relationships, highly heterogeneous populations, multiple product offeri...
The gustatory, olfactory, and trigeminal systems are anatomically separated. However, they interact cognitively to give rise to oral perception, which can significantly affect health and quality of life. We built a Supervised Learning (SL) regression model that, exploiting participants’ features, was capable of automatically analyzing with high pre...
The k-nearest neighbors (KNN) algorithm has been widely used for classification analysis in machine learning. However, it suffers from noise samples that reduce its classification ability and therefore prediction accuracy. This article introduces the high-level k-nearest neighbors (HLKNN) method, a new technique for enhancing the k-nearest neighbor...
Citations
... The efficacy of machine-learning method for optical correction surpasses that of the standard response matrix approach. Subsequent research has explored the application of both local and global corrections [1,2]. At the Stanford Positron Electron Asymmetric Ring, an adaptive feedback technique is employed to swiftly and consistently optimize magnets to reduce betatron oscillations [3]. ...
Machine learning (ML) has become a valuable tool in particle accelerator control, with growing potential for beam parameter correction. In this study, we present preliminary ML applications at HLS-II, using Lasso regression for online tune correction and a neural network (NN) for beta function simulation correction. Both models were trained with supervised learning on measured beam parameter data, while an improved genetic algorithm optimized the NN structure. Our results show that the ML-based approach achieves competitive correction quality with fewer steps, making it a promising method for future particle accelerator applications and other fields.
... An effort has also been made toward developing ML-based accelerator controllers using Bayesian and GP approaches for accelerator tuning [275], [288], [289], [290], [291], [292], including various applications at the Large Hadron Collider for optics corrections and detecting faulty beam position monitors [293], [294], [295], [296], and PC expansion-based surrogate models for UQ [297]. RL tools have also been developed for online accelerator optimization [298], [299], [300], [301]. ...
Data-driven science and technology offer transformative tools and methods to science. This review article highlights the latest development and progress in the interdisciplinary field of data-driven plasma science (DDPS), i.e., plasma science whose progress is driven strongly by data and data analyses. Plasma is considered to be the most ubiquitous form of observable matter in the universe. Data associated with plasmas can, therefore, cover extremely large spatial and temporal scales, and often provide essential information for other scientific disciplines. Thanks to the latest technological developments, plasma experiments, observations, and computation now produce a large amount of data that can no longer be analyzed or interpreted manually. This trend now necessitates a highly sophisticated use of high-performance computers for data analyses, making artificial intelligence and machine learning vital components of DDPS. This article contains seven primary sections, in addition to the introduction and summary. Following an overview of fundamental data-driven science, five other sections cover widely studied topics of plasma science and technologies, i.e., basic plasma physics and laboratory experiments, magnetic confinement fusion, inertial confinement fusion and high-energy-density physics, space and astronomical plasmas, and plasma technologies for industrial and other applications. The final section before the summary discusses plasma-related databases that could significantly contribute to DDPS. Each primary section starts with a brief introduction to the topic, discusses the state-of-the-art developments in the use of data and/or data-scientific approaches, and presents the summary and outlook. Despite the recent impressive signs of progress, the DDPS is still in its infancy. This article attempts to offer a broad perspective on the development of this field and identify where further innovations are required.
... Neural networks (NN) are powerful machine learning (ML) tools that can extract complex physical relationships directly from data and have been used for speeding up the studies of complex physical systems. [37][38][39][40][41][42][43][44] Incredibly powerful and flexible physicsinformed neural networks (PINNs), which include soft constraints in the NN's cost function, have been developed and have shown great capabilities for complex fluid dynamics simulations, 45 material science, 46 for symplectic single particle tracking, 47 for learning molecular force fields, 48 and for large classes of partial differential equations. 49,50 For the problem of mapping current density J to an estimateB, the PINN approach is to train a neural network with a cost function defined as ...
We present a physics-constrained neural network (PCNN) approach to solving Maxwell’s equations for the electromagnetic fields of intense relativistic charged particle beams. We create a 3D convolutional PCNN to map time-varying current and charge densities J(r, t) and ρ(r, t) to vector and scalar potentials A(r, t) and φ(r, t) from which we generate electromagnetic fields according to Maxwell’s equations: B = ∇ × A and E = −∇ φ − ∂A/ ∂t. Our PCNNs satisfy hard constraints, such as ∇ · B = 0, by construction. Soft constraints push A and φ toward satisfying the Lorenz gauge.
... For example, for the problem of mapping current density J to an estimate Bˆ of the associated magnetic field B we build Eq. 2 into the structure of our NN and generate the vector potential Aˆ , which defines the magnetic field as K = ∇ × K ⟹ ∇ ⋅ K = ∇ ⋅ N∇ × K O = 0, (6) which satisfies the physics constraint by construction. Neural networks (NN) are powerful machine learning (ML) tools which can extract complex physical relationships directly from data and have been used for speeding up the studies of complex physical systems [36][37][38][39][40][41][42][43]. Incredibly powerful and flexible physics-informed neural networks (PINNs), which include soft constraints in the NN's cost function, have been developed and have shown great capabilities for complex fluid dynamics simulations [44], material science [45], for symplectic single particle tracking [46], for learning molecular force fields [47], and for large classes of partial differential equations [48][49][50]. ...
We present a physics-constrained neural network (PCNN) approach to solving Maxwell's equations for the electromagnetic fields of intense relativistic charged particle beams. We create a 3D convolutional PCNN to map time-varying current and charge densities J(r,t) and p(r,t) to vector and scalar potentials A(r,t) and V(r,t) from which we generate electromagnetic fields according to Maxwell's equations: B=curl(A), E=-div(V)-dA/dt. Our PCNNs satisfy hard constraints, such as div(B)=0, by construction. Soft constraints push A and V towards satisfying the Lorenz gauge.
... Contributions to this issue represent and attempt to cover a broad domain of topics in beam physics, with a loose classification into three categories of advances, challenges, and applications. References [13][14][15][16][17][18][19][20][21] provide a global overview of the applications of advanced beam dynamics concepts to existing or future machines. The nonlinear beam dynamics combined with the use of electron lenses has been considered in the framework of the High-Luminosity LHC Project and discussed in [21]. ...
... A combined-function optics has been proposed as an alternative to the standard separate approach paradigm for high-energy colliders [19] with the goal of increasing the dipole filling factor. Machine learning entered accelerator physics, and applications to measurements of optical parameters at the LHC are reported in [18]. An interesting application to future muon colliders of a special optics concept based on skew quadrupoles is considered in [17]. ...
... The development of ML-based tools for particle accelerator applications is an active area of research. At CERN, supervised learning techniques are being applied for the reconstruction of magnet errors in the incredibly large (thousands of magnets) LHC lattice [4]. At the LCLS, Bayesian methods have been developed for online accelerator tuning [5], Bayesian methods with safety constraints are being developed at the SwissFEL and the High-Intensity Proton Accelerator at PSI [6], at SLAC Bayesian methods are being developed for the challenging problem of hysteresis [7] and surrogate models are being developed for the beam at the injector [8], and at LANL researchers have been developing methods to combine neural networks with modelindependent adaptive feedback for automatic control of the * ascheink@lanl.gov ...
Particle accelerators are complicated machines with thousands of coupled time varying components. The electromagnetic fields of accelerator devices such as magnets and RF cavities drift and are uncertain due to external disturbances, vibrations, temperature changes, and hysteresis. Accelerated charged particle beams are complex objects with 6D phase space dynamics governed by collective effects such as space charge forces, coherent synchrotron radiation, and whose initial phase space distributions change in unexpected and difficult to measure ways. This two-part tutorial presents recent developments in Bayesian methods and adaptive machine learning (ML) techniques for accelerators. Part 1: We introduce Bayesian control algorithms, and we describe how these algorithms can be customized to solve practical accelerator specific problems, including online characterization and optimization. Part 2: We give an overview of adaptive ML (AML) combining adaptive model-independent feedback within physics-informed ML architectures to make ML tools robust to time-variation (distribution shift) and to enable their use further beyond the span of the training data without relying on retraining.
... An effort has also been made towards developing MLbased accelerator controllers using Bayesian and Gaussian Process (GP) approaches for accelerator tuning [260][261][262][263][264][265], including various applications at the Large Hadron Collider for optics corrections and detecting faulty beam position monitors [266][267][268][269], and polynomial chaos expansion-based surrogate models for uncertainty quantification [270]. Reinforcement learning (RL) tools have also been developed for online accelerator optimization [271][272][273][274]. ...
Data science and technology offer transformative tools and methods to science. This review article highlights latest development and progress in the interdisciplinary field of data-driven plasma science (DDPS). A large amount of data and machine learning algorithms go hand in hand. Most plasma data, whether experimental, observational or computational, are generated or collected by machines today. It is now becoming impractical for humans to analyze all the data manually. Therefore, it is imperative to train machines to analyze and interpret (eventually) such data as intelligently as humans but far more efficiently in quantity. Despite the recent impressive progress in applications of data science to plasma science and technology, the emerging field of DDPS is still in its infancy. Fueled by some of the most challenging problems such as fusion energy, plasma processing of materials, and fundamental understanding of the universe through observable plasma phenomena, it is expected that DDPS continues to benefit significantly from the interdisciplinary marriage between plasma science and data science into the foreseeable future.
... Researchers at SLAC National Accelerator Laboratory have also developed neural network-based virtual LPS diagnostics [33], and methods which utilize not only accelerator parameters as inputs, but also spectral measurements for increased resolution and prediction accuracy [34]. At CERN ML tools have also been developed as virtual diagnostics for not just the accelerated beam, but for the accelerator itself, for example for identifying magnet errors based on beam measurements [35]. At CERN surrogate models have also been developed for fast simulation studies of the CLIC final focus system, mapping sextupole offsets to luminosity and beam sizes without requiring computationally expensive tracking and beam-beam simulations [36]. ...
Particle accelerators for high energy physics will generate TeV-scale particle beams in large, multi-Km size machines colliding high brightness beams at the interaction point [1-4]. The high luminosity in such machines is achieved by producing very small asymmetric beam size at the interaction point, with short durations to minimize beam-beam effects. Tuning energy, timing and position of the beam for optimal performance will require high-precision controls of amplitude and phase of high-frequency electromagnetic fields and real-time processing of complex algorithms. The stability of the colliding beams has a large impact on the collider's effective luminosity. Therefore, the technology readiness level of diagnostic and control systems will be a major consideration in the collider design. The technical requirements of such systems depend on the specifics of beam parameters, such as transverse and longitudinal dimensions, charge/pulse and beam pulse format, which are driven by the accelerating technology of choice. While feedback systems with single bunch position monitor resolution below 50 nm and latency <300 ns have been demonstrated in beam test facilities, many advanced collider concepts make use of higher repetition rates, brighter beams and higher accelerating frequencies, and will require better performance, up to 1-2 order of magnitude, demanding aggressive R&D to be able to deliver and maintain the targeted luminosity.
... According to the relationship between the Beam Position Monitor position and the quadrupole, the effect of optical correction is better than that of the traditional response matrix approach, and the application of local and global corrections was completed in a later study. In addition, self-encoder neural networks and linear regression measurement data denoising and reconstruction techniques have been proposed [5,6]. At the Stanford Positron Electron Asymmetric Ring, an adaptive feedback method to rapidly optimize and continuously optimize the magnet, even for an unknown magnet that changes rapidly, can continuously adjust all other magnets to minimize mismatches and the resulting betatron oscillations [7]. ...
The theory of tune feedback correction and the principle of a feedback algorithm based on machine learning are introduced, with a focus on the application of lasso regression for tune feedback correction. Simulation verification and online feedback correction results are presented. The results show that, after applying machine learning, the feedback accuracy of the tune feedback system was higher, and the betatron tune stability was further improved.
... More than this, they can boost the quality of optics corrections, as they provide insight into the sources of local errors in the accelerator. Some novel findings are presented in [30] and cover cases where the data are from numerical simulations and beam measurements. ...
Particle accelerators are among the most complex instruments conceived by physicists for the exploration of the fundamental laws of nature. Of relevance for particle physics are the high-energy colliders, such as the CERN Large Hadron Collider (LHC), which hosts particle physics experiments that are probing the Standard Model predictions and looking for signs of physics beyond the standard model.