-
Parallelizing non-linear sequential models over the sequence length
Authors:
Yi Heng Lim,
Qi Zhu,
Joshua Selfridge,
Muhammad Firmansyah Kasim
Abstract:
Sequential models, such as Recurrent Neural Networks and Neural Ordinary Differential Equations, have long suffered from slow training due to their inherent sequential nature. For many years this bottleneck has persisted, as many thought sequential models could not be parallelized. We challenge this long-held belief with our parallel algorithm that accelerates GPU evaluation of sequential models b…
▽ More
Sequential models, such as Recurrent Neural Networks and Neural Ordinary Differential Equations, have long suffered from slow training due to their inherent sequential nature. For many years this bottleneck has persisted, as many thought sequential models could not be parallelized. We challenge this long-held belief with our parallel algorithm that accelerates GPU evaluation of sequential models by up to 3 orders of magnitude faster without compromising output accuracy. The algorithm does not need any special structure in the sequential models' architecture, making it applicable to a wide range of architectures. Using our method, training sequential models can be more than 10 times faster than the common sequential method without any meaningful difference in the training results. Leveraging this accelerated training, we discovered the efficacy of the Gated Recurrent Unit in a long time series classification problem with 17k time samples. By overcoming the training bottleneck, our work serves as the first step to unlock the potential of non-linear sequential models for long sequence problems.
△ Less
Submitted 16 January, 2024; v1 submitted 21 September, 2023;
originally announced September 2023.
-
Investigating Mechanisms of State Localization in Highly-Ionized Dense Plasmas
Authors:
Thomas Gawne,
Thomas Campbell,
Alessandro Forte,
Patrick Hollebon,
Gabriel Perez-Callejo,
Oliver Humphries,
Oliver Karnbach,
Muhammad F. Kasim,
Thomas R. Preston,
Hae Ja Lee,
Alan Miscampbell,
Quincy Y. van den Berg,
Bob Nagler,
Shenyuan Ren,
Ryan B. Royle,
Justin S. Wark,
Sam M. Vinko
Abstract:
We present the first experimental observation of K$_β$ emission from highly charged Mg ions at solid density, driven by intense x-rays from a free electron laser. The presence of K$_β$ emission indicates the $n=3$ atomic shell is relocalized for high charge states, providing an upper constraint on the depression of the ionization potential. We explore the process of state relocalization in dense p…
▽ More
We present the first experimental observation of K$_β$ emission from highly charged Mg ions at solid density, driven by intense x-rays from a free electron laser. The presence of K$_β$ emission indicates the $n=3$ atomic shell is relocalized for high charge states, providing an upper constraint on the depression of the ionization potential. We explore the process of state relocalization in dense plasmas from first principles using finite-temperature density functional theory alongside a wavefunction localization metric, and find excellent agreement with experimental results.
△ Less
Submitted 14 August, 2023; v1 submitted 8 February, 2023;
originally announced February 2023.
-
Constants of motion network
Authors:
Muhammad Firmansyah Kasim,
Yi Heng Lim
Abstract:
The beauty of physics is that there is usually a conserved quantity in an always-changing system, known as the constant of motion. Finding the constant of motion is important in understanding the dynamics of the system, but typically requires mathematical proficiency and manual analytical work. In this paper, we present a neural network that can simultaneously learn the dynamics of the system and…
▽ More
The beauty of physics is that there is usually a conserved quantity in an always-changing system, known as the constant of motion. Finding the constant of motion is important in understanding the dynamics of the system, but typically requires mathematical proficiency and manual analytical work. In this paper, we present a neural network that can simultaneously learn the dynamics of the system and the constants of motion from data. By exploiting the discovered constants of motion, it can produce better predictions on dynamics and can work on a wider range of systems than Hamiltonian-based neural networks. In addition, the training progresses of our method can be used as an indication of the number of constants of motion in a system which could be useful in studying a novel physical system.
△ Less
Submitted 4 October, 2022; v1 submitted 22 August, 2022;
originally announced August 2022.
-
Unifying physical systems' inductive biases in neural ODE using dynamics constraints
Authors:
Yi Heng Lim,
Muhammad Firmansyah Kasim
Abstract:
Conservation of energy is at the core of many physical phenomena and dynamical systems. There have been a significant number of works in the past few years aimed at predicting the trajectory of motion of dynamical systems using neural networks while adhering to the law of conservation of energy. Most of these works are inspired by classical mechanics such as Hamiltonian and Lagrangian mechanics as…
▽ More
Conservation of energy is at the core of many physical phenomena and dynamical systems. There have been a significant number of works in the past few years aimed at predicting the trajectory of motion of dynamical systems using neural networks while adhering to the law of conservation of energy. Most of these works are inspired by classical mechanics such as Hamiltonian and Lagrangian mechanics as well as Neural Ordinary Differential Equations. While these works have been shown to work well in specific domains respectively, there is a lack of a unifying method that is more generally applicable without requiring significant changes to the neural network architectures. In this work, we aim to address this issue by providing a simple method that could be applied to not just energy-conserving systems, but also dissipative systems, by including a different inductive bias in different cases in the form of a regularisation term in the loss function. The proposed method does not require changing the neural network architecture and could form the basis to validate a novel idea, therefore showing promises to accelerate research in this direction.
△ Less
Submitted 3 August, 2022;
originally announced August 2022.
-
Reducing the Long Tail Losses in Scientific Emulations with Active Learning
Authors:
Yi Heng Lim,
Muhammad Firmansyah Kasim
Abstract:
Deep-learning-based models are increasingly used to emulate scientific simulations to accelerate scientific research. However, accurate, supervised deep learning models require huge amount of labelled data, and that often becomes the bottleneck in employing neural networks. In this work, we leveraged an active learning approach called core-set selection to actively select data, per a pre-defined b…
▽ More
Deep-learning-based models are increasingly used to emulate scientific simulations to accelerate scientific research. However, accurate, supervised deep learning models require huge amount of labelled data, and that often becomes the bottleneck in employing neural networks. In this work, we leveraged an active learning approach called core-set selection to actively select data, per a pre-defined budget, to be labelled for training. To further improve the model performance and reduce the training costs, we also warm started the training using a shrink-and-perturb trick. We tested on two case studies in different fields, namely galaxy halo occupation distribution modelling in astrophysics and x-ray emission spectroscopy in plasma physics, and the results are promising: we achieved competitive overall performance compared to using a random sampling baseline, and more importantly, successfully reduced the larger absolute losses, i.e. the long tail in the loss distribution, at virtually no overhead costs.
△ Less
Submitted 9 January, 2022; v1 submitted 15 November, 2021;
originally announced November 2021.
-
DQC: a Python program package for Differentiable Quantum Chemistry
Authors:
Muhammad F. Kasim,
Susi Lehtola,
Sam M. Vinko
Abstract:
Automatic differentiation represents a paradigm shift in scientific programming, where evaluating both functions and their derivatives is required for most applications. By removing the need to explicitly derive expressions for gradients, development times can be be shortened, and calculations simplified. For these reasons, automatic differentiation has fueled the rapid growth of a variety of soph…
▽ More
Automatic differentiation represents a paradigm shift in scientific programming, where evaluating both functions and their derivatives is required for most applications. By removing the need to explicitly derive expressions for gradients, development times can be be shortened, and calculations simplified. For these reasons, automatic differentiation has fueled the rapid growth of a variety of sophisticated machine learning techniques over the past decade, but is now also increasingly showing its value to support {\it ab initio} simulations of quantum systems, and enhance computational quantum chemistry. Here we present an open-source differentiable quantum chemistry simulation code, DQC, and explore applications facilitated by automatic differentiation: (1) calculating molecular perturbation properties; (2) reoptimizing a basis set for hydrocarbons; (3) checking the stability of self-consistent field wave functions; and (4) predicting molecular properties via alchemical perturbations.
△ Less
Submitted 22 October, 2021;
originally announced October 2021.
-
Failure modes and downtime of radiotherapy linear accelerators and Multi-Leaf Collimators in Indonesia
Authors:
Gregory S. Peiris,
Supriyanto A. Pawiro,
Muhammad F. Kasim,
Suzanne L. Sheehy
Abstract:
The lack of equitable access to radiotherapy linear accelerators (LINACs) is a substantial barrier to cancer care in Low and Middle-Income Countries (LMICs). Aside from the issue of cost, there are also issues of robustness of state-of-the-art LINACs in LMICs, which are subject to mechanical and electrical breakdowns, which can result in downtimes ranging from days to months. Recent research has i…
▽ More
The lack of equitable access to radiotherapy linear accelerators (LINACs) is a substantial barrier to cancer care in Low and Middle-Income Countries (LMICs). Aside from the issue of cost, there are also issues of robustness of state-of-the-art LINACs in LMICs, which are subject to mechanical and electrical breakdowns, which can result in downtimes ranging from days to months. Recent research has identified disparities in failure frequency and downtimes between LMICs (Nigeria, Botswana) and a High-Income Countries (HIC, the UK), and highlighted the need for additional data and study particularly relating to Multi-Leaf Collimators (MLCs). This study analyses data from 14 Indonesian Hospitals with a total of 19 LINACs and shows the pathways to failure of radiotherapy LINACs and frequency of breakdowns with an additional focus on the Multi-Leaf Collimator (MLC) subsystem. We found that LINACs throughout Indonesia are out of operation for 7 times longer than HICs and the Mean Time Between Failures of a LINAC in Indonesia is 341.58 Hours, or about 14 days. Furthermore, of the LINACs with an MLC fitted, $59.02^{+1.98}_{-1.61}$% of mechanical faults are due to the MLC and $57.14^{+0.78}_{-1.27}$% of cases requiring a replacement component are related to the MLC. These results outline the need to reassess the current generation of RT LINACs and ultimately work towards guiding future designs to be robust enough for all environments.
△ Less
Submitted 7 November, 2021; v1 submitted 18 May, 2021;
originally announced May 2021.
-
Learning the exchange-correlation functional from nature with fully differentiable density functional theory
Authors:
Muhammad F. Kasim,
Sam M. Vinko
Abstract:
Improving the predictive capability of molecular properties in ab initio simulations is essential for advanced material discovery. Despite recent progress making use of machine learning, utilizing deep neural networks to improve quantum chemistry modelling remains severely limited by the scarcity and heterogeneity of appropriate experimental data. Here we show how training a neural network to repl…
▽ More
Improving the predictive capability of molecular properties in ab initio simulations is essential for advanced material discovery. Despite recent progress making use of machine learning, utilizing deep neural networks to improve quantum chemistry modelling remains severely limited by the scarcity and heterogeneity of appropriate experimental data. Here we show how training a neural network to replace the exchange-correlation functional within a fully-differentiable three-dimensional Kohn-Sham density functional theory (DFT) framework can greatly improve simulation accuracy. Using only eight experimental data points on diatomic molecules, our trained exchange-correlation networks enable improved prediction accuracy of atomization energies across a collection of 104 molecules containing new bonds and atoms that are not present in the training dataset.
△ Less
Submitted 16 June, 2021; v1 submitted 8 February, 2021;
originally announced February 2021.
-
Derivatives of partial eigendecomposition of a real symmetric matrix for degenerate cases
Authors:
Muhammad Firmansyah Kasim
Abstract:
This paper presents the forward and backward derivatives of partial eigendecomposition, i.e. where it only obtains some of the eigenpairs, of a real symmetric matrix for degenerate cases. The numerical calculation of forward and backward derivatives can be implemented even if the degeneracy never disappears and only some eigenpairs are available.
This paper presents the forward and backward derivatives of partial eigendecomposition, i.e. where it only obtains some of the eigenpairs, of a real symmetric matrix for degenerate cases. The numerical calculation of forward and backward derivatives can be implemented even if the degeneracy never disappears and only some eigenpairs are available.
△ Less
Submitted 9 November, 2020;
originally announced November 2020.
-
$ξ$-torch: differentiable scientific computing library
Authors:
Muhammad F. Kasim,
Sam M. Vinko
Abstract:
Physics-informed learning has shown to have a better generalization than learning without physical priors.
However, training physics-informed deep neural networks requires some aspect of physical simulations to be written in a differentiable manner.
Unfortunately, some operations and functionals commonly used in physical simulations are scattered, hard to integrate, and lack higher order deriv…
▽ More
Physics-informed learning has shown to have a better generalization than learning without physical priors.
However, training physics-informed deep neural networks requires some aspect of physical simulations to be written in a differentiable manner.
Unfortunately, some operations and functionals commonly used in physical simulations are scattered, hard to integrate, and lack higher order derivatives which are needed in physical simulations.
In this work, we present $ξ$-torch, a library of differentiable functionals for scientific simulations.
Example functionals are a root finder and an initial value problem solver, among others.
The gradient of functionals in $ξ$-torch are written based on their analytical expression to improve numerical stability and reduce memory requirements.
$ξ$-torch also provides second and higher order derivatives of the functionals which are rarely available in existing packages.
We show two applications of this library in optimizing parameters in physics simulations.
The library and all test cases in this work can be found at https://github.com/xitorch/xitorch/ and the documentation at https://xitorch.readthedocs.io.
△ Less
Submitted 5 October, 2020;
originally announced October 2020.
-
How many can you infect? Simple (and naive) methods of estimating the reproduction number
Authors:
H. Susanto,
V. R. Tjahjono,
A. Hasan,
M. F. Kasim,
N. Nuraini,
E. R. M. Putri,
R. Kusdiantara,
H. Kurniawan
Abstract:
This is a pedagogical paper on estimating the number of people that can be infected by one infectious person during an epidemic outbreak, known as the reproduction number. Knowing the number is crucial for develo** policy responses. There are generally two types of such a number, i.e., basic and effective (or instantaneous). While basic reproduction number is the average expected number of cases…
▽ More
This is a pedagogical paper on estimating the number of people that can be infected by one infectious person during an epidemic outbreak, known as the reproduction number. Knowing the number is crucial for develo** policy responses. There are generally two types of such a number, i.e., basic and effective (or instantaneous). While basic reproduction number is the average expected number of cases directly generated by one case in a population where all individuals are susceptible, effective reproduction number is the number of cases generated in the current state of a population. In this paper, we exploit the deterministic susceptible-infected-removed (SIR) model to estimate them through three different numerical approximations. We apply the methods to the pandemic COVID-19 in Italy to provide insights into the spread of the disease in the country. We see that the effect of the national lockdown in slowing down the disease exponential growth appeared about two weeks after the implementation date. We also discuss available improvements to the simple (and naive) methods that have been made by researchers in the field.
Authors of this paper are members of the SimcovID (Simulasi dan Pemodelan COVID-19 Indonesia) collaboration.
△ Less
Submitted 28 June, 2020;
originally announced June 2020.
-
Building high accuracy emulators for scientific simulations with deep neural architecture search
Authors:
M. F. Kasim,
D. Watson-Parris,
L. Deaconu,
S. Oliver,
P. Hatfield,
D. H. Froula,
G. Gregori,
M. Jarvis,
S. Khatiwala,
J. Korenaga,
J. Topp-Mugglestone,
E. Viezzer,
S. M. Vinko
Abstract:
Computer simulations are invaluable tools for scientific discovery. However, accurate simulations are often slow to execute, which limits their applicability to extensive parameter exploration, large-scale data analysis, and uncertainty quantification. A promising route to accelerate simulations by building fast emulators with machine learning requires large training datasets, which can be prohibi…
▽ More
Computer simulations are invaluable tools for scientific discovery. However, accurate simulations are often slow to execute, which limits their applicability to extensive parameter exploration, large-scale data analysis, and uncertainty quantification. A promising route to accelerate simulations by building fast emulators with machine learning requires large training datasets, which can be prohibitively expensive to obtain with slow simulations. Here we present a method based on neural architecture search to build accurate emulators even with a limited number of training data. The method successfully accelerates simulations by up to 2 billion times in 10 scientific cases including astrophysics, climate science, biogeochemistry, high energy density physics, fusion energy, and seismology, using the same super-architecture, algorithm, and hyperparameters. Our approach also inherently provides emulator uncertainty estimation, adding further confidence in their use. We anticipate this work will accelerate research involving expensive simulations, allow more extensive parameters exploration, and enable new, previously unfeasible computational discovery.
△ Less
Submitted 8 October, 2020; v1 submitted 17 January, 2020;
originally announced January 2020.
-
Time-resolved XUV Opacity Measurements of Warm-Dense Aluminium
Authors:
S. M. Vinko,
V. Vozda,
J. Andreasson,
S. Bajt,
J. Bielecki,
T. Burian,
J. Chalupsky,
O. Ciricosta,
M. P. Desjarlais,
H. Fleckenstein,
J. Hajdu,
V. Hajkova,
P. Hollebon,
L. Juha,
M. F. Kasim,
E. E. McBride,
K. Muehlig,
T. R. Preston,
D. S. Rackstraw,
S. Roling,
S. Toleikis,
J. S. Wark,
H. Zacharias
Abstract:
The free-free opacity in plasmas is fundamental to our understanding of energy transport in stellar interiors and for inertial confinement fusion research. However, theoretical predictions in the challenging dense plasma regime are conflicting and there is a dearth of accurate experimental data to allow for direct model validation. Here we present time-resolved transmission measurements in solid-d…
▽ More
The free-free opacity in plasmas is fundamental to our understanding of energy transport in stellar interiors and for inertial confinement fusion research. However, theoretical predictions in the challenging dense plasma regime are conflicting and there is a dearth of accurate experimental data to allow for direct model validation. Here we present time-resolved transmission measurements in solid-density Al heated by an XUV free-electron laser. We use a novel functional optimization approach to extract the temperature-dependent absorption coefficient directly from an oversampled pool of single-shot measurements, and find a pronounced enhancement of the opacity as the plasma is heated to temperatures of order the Fermi energy. Plasma heating and opacity-enhancement is observed on ultrafast time scales, within the duration of the femtosecond XUV pulse. We attribute further rises in the opacity on ps timescales to melt and the formation of warm-dense matter.
△ Less
Submitted 18 January, 2020;
originally announced January 2020.
-
Map** the Electronic Structure of Warm Dense Nickel via Resonant Inelastic X-ray Scattering
Authors:
O. S. Humphries,
R. S. Marjoribanks,
Q. van den Berg,
E. C. Galtier,
M. F. Kasim,
H. J. Lee,
A. J. F. Miscampbell,
B. Nagler,
R. Royle,
J. S. Wark,
S. M. Vinko
Abstract:
The development of high-brightness free-electron lasers (FEL) has revolutionised our ability to create and study matter in the high-energy-density (HED) regime. Current diagnostic techniques have been very successful in yielding information on fundamental thermodynamic plasma properties, but provide only limited or indirect information on the detailed quantum structure of these systems, and on how…
▽ More
The development of high-brightness free-electron lasers (FEL) has revolutionised our ability to create and study matter in the high-energy-density (HED) regime. Current diagnostic techniques have been very successful in yielding information on fundamental thermodynamic plasma properties, but provide only limited or indirect information on the detailed quantum structure of these systems, and on how it is affected by ionization dynamics. Here we show how the electronic structure of solid-density nickel, heated to temperatures of 10's of eV on femtosecond timescales, can be studied by resonant (Raman) inelastic x-ray scattering (RIXS) using the Linac Coherent Light Source FEL. We present single-shot measurements of the valence density of states in the x-ray-heated transient system, and extract simultaneously electron temperatures, ionization, and ionization potential energies. The RIXS spectrum provides a wealth of information on the valence structure of the HED system that goes beyond what can be extracted from x-ray absorption or emission spectroscopy alone.
△ Less
Submitted 26 October, 2020; v1 submitted 16 January, 2020;
originally announced January 2020.
-
Efficient Parameter Sampling for Neural Network Construction
Authors:
Drimik Roy Chowdhury,
Muhammad Firmansyah Kasim
Abstract:
The customizable nature of deep learning models have allowed them to be successful predictors in various disciplines. These models are often trained with respect to thousands or millions of instances for complicated problems, but the gathering of such an immense collection may be infeasible and expensive. However, what often occurs is the pollution of redundant information from these instances to…
▽ More
The customizable nature of deep learning models have allowed them to be successful predictors in various disciplines. These models are often trained with respect to thousands or millions of instances for complicated problems, but the gathering of such an immense collection may be infeasible and expensive. However, what often occurs is the pollution of redundant information from these instances to the deep learning models. This paper outlines an algorithm that dynamically selects and appends instances to a training dataset from uncertain regions of the parameter space based on differences in predictions from multiple convolutional neural networks (CNNs). These CNNs are also simultaneously trained on this growing dataset to construct more accurate and knowledgable models. The methodology presented has reduced training dataset sizes by almost 90% and maintained predictive power in two diagnostics of high energy density physics.
△ Less
Submitted 22 December, 2019;
originally announced December 2019.
-
Retrieving fields from proton radiography without source profiles
Authors:
M. F. Kasim,
A. F. A. Bott,
P. Tzeferacos,
D. Q. Lamb,
G. Gregori,
S. M. Vinko
Abstract:
Proton radiography is a technique in high energy density science to diagnose magnetic and/or electric fields in a plasma by firing a proton beam and detecting its modulated intensity profile on a screen. Current approaches to retrieve the integrated field from the modulated intensity profile require the unmodulated beam intensity profile before the interaction, which is rarely available experiment…
▽ More
Proton radiography is a technique in high energy density science to diagnose magnetic and/or electric fields in a plasma by firing a proton beam and detecting its modulated intensity profile on a screen. Current approaches to retrieve the integrated field from the modulated intensity profile require the unmodulated beam intensity profile before the interaction, which is rarely available experimentally due to shot-to-shot variability. In this paper, we present a statistical method to retrieve the integrated field without needing to know the exact source profile. We apply our method to experimental data, showing the robustness of our approach. Our proposed technique allows not only for the retrieval of the path-integrated fields, but also of the statistical properties of the fields.
△ Less
Submitted 13 September, 2019; v1 submitted 30 May, 2019;
originally announced May 2019.
-
Inverse Problem Instabilities in Large-Scale Plasma Modelling
Authors:
M. F. Kasim,
T. P. Galligan,
J. Topp-Mugglestone,
G. Gregori,
S. M. Vinko
Abstract:
Our understanding of physical systems generally depends on our ability to match complex computational modelling with measured experimental outcomes. However, simulations with large parameter spaces suffer from inverse problem instabilities, where similar simulated outputs can map back to very different sets of input parameters. While of fundamental importance, such instabilities are seldom resolve…
▽ More
Our understanding of physical systems generally depends on our ability to match complex computational modelling with measured experimental outcomes. However, simulations with large parameter spaces suffer from inverse problem instabilities, where similar simulated outputs can map back to very different sets of input parameters. While of fundamental importance, such instabilities are seldom resolved due to the intractably large number of simulations required to comprehensively explore parameter space. Here we show how Bayesian machine learning can be used to address inverse problem instabilities, and apply it to two popular experimental diagnostics in plasma physics. We find that the extraction of information from measurements simply on the basis of agreement with simulations is unreliable, and leads to a significant underestimation of uncertainties. We describe how to statistically quantify the effect of unstable inverse models, and describe an approach to experimental design that mitigates its impact.
△ Less
Submitted 3 July, 2018; v1 submitted 21 May, 2018;
originally announced May 2018.
-
Validating Continuum Lowering Models via Multi-Wavelength Measurements of Integrated X-ray Emission
Authors:
M. F. Kasim,
J. S. Wark,
S. M. Vinko
Abstract:
X-ray emission spectroscopy is a well-established technique used to study continuum lowering in dense plasmas. It relies on accurate atomic physics models to robustly reproduce high-resolution emission spectra, and depends on our ability to identify spectroscopic signatures such as emission lines or ionization edges of individual charge states within the plasma. Here we describe a method that forg…
▽ More
X-ray emission spectroscopy is a well-established technique used to study continuum lowering in dense plasmas. It relies on accurate atomic physics models to robustly reproduce high-resolution emission spectra, and depends on our ability to identify spectroscopic signatures such as emission lines or ionization edges of individual charge states within the plasma. Here we describe a method that forgoes these requirements, enabling the validation of different continuum lowering models based solely on the total intensity of plasma emission in systems driven by narrow-bandwidth x-ray pulses across a range of wavelengths. The method is tested on published Al spectroscopy data and applied to the new case of solid-density partially-ionized Fe plasmas, where extracting ionization edges directly is precluded by the significant overlap of emission from a wide range of charge states.
△ Less
Submitted 4 February, 2018;
originally announced February 2018.
-
Three dimensional spectrometer
Authors:
Muhammad Firmansyah Kasim,
Peter Norreys
Abstract:
We present a novel design of 3D spectrometer that can retrieve 3D spectral profile in a single measurement. The 3D spectrometer design is built upon the concept of compressed sensing to make it possible to retrieve 3D information from 2D data from a screen/camera. In contrast to common spectrometers, the 3D spectrometer uses a wide slit instead of a narrow slit and retrieve the 3D datacube that co…
▽ More
We present a novel design of 3D spectrometer that can retrieve 3D spectral profile in a single measurement. The 3D spectrometer design is built upon the concept of compressed sensing to make it possible to retrieve 3D information from 2D data from a screen/camera. In contrast to common spectrometers, the 3D spectrometer uses a wide slit instead of a narrow slit and retrieve the 3D datacube that consists of 2D spatial and 1D spectral information. Numerical tests were performed to simulate the retrieval of spectral profiles. The results show that the retrieved profiles match well the original profiles. It is also shown that the retrieved signal from the 3D spectrometer is robust enough for a further post-processing analysis.
△ Less
Submitted 30 January, 2018;
originally announced February 2018.
-
Infinite dimensional optimistic optimisation with applications on physical systems
Authors:
Muhammad F. Kasim,
Peter A. Norreys
Abstract:
This paper presents a novel numerical optimisation method for infinite dimensional optimisation. The functional optimisation makes minimal assumptions about the functional and without any specific knowledge on the derivative of the functional. The algorithm has been tested on several physical systems (brachistochrone and catenary problems) and it is shown that the solutions obtained are close to t…
▽ More
This paper presents a novel numerical optimisation method for infinite dimensional optimisation. The functional optimisation makes minimal assumptions about the functional and without any specific knowledge on the derivative of the functional. The algorithm has been tested on several physical systems (brachistochrone and catenary problems) and it is shown that the solutions obtained are close to the actual solutions in one thousand functional evaluations. It is also shown that for the tested cases, the new algorithm provides better convergence to the optimum value compared to the tested existing algorithms.
△ Less
Submitted 17 November, 2016;
originally announced November 2016.
-
Essential criteria for efficient pulse amplification via Raman and Brillouin scattering
Authors:
R. M. G. M. Trines,
E. P. Alves,
E. Webb,
J. Vieira,
F. Fiuza,
R. A. Fonseca,
L. O. Silva,
J. Sadler,
N. Ratan,
L. Ceurvorst,
M. F. Kasim,
M. Tabak,
D. Froula,
D. Haberberger,
P. A. Norreys,
R. A. Cairns,
R. Bingham
Abstract:
Raman and Brillouin amplification are two schemes for amplifying and compressing short laser pulses in plasma. Analytical models have already been derived for both schemes, but the full consequences of these models are little known or used. Here, we present new criteria that govern the evolution of the attractor solution for the seed pulse in Raman and Brillouin amplification, and show how the ini…
▽ More
Raman and Brillouin amplification are two schemes for amplifying and compressing short laser pulses in plasma. Analytical models have already been derived for both schemes, but the full consequences of these models are little known or used. Here, we present new criteria that govern the evolution of the attractor solution for the seed pulse in Raman and Brillouin amplification, and show how the initial laser pulses need to be shaped to control the properties of the final amplified seed and improve the amplification efficiency.
△ Less
Submitted 14 November, 2016;
originally announced November 2016.
-
Machine learning applied to proton radiography
Authors:
Nicholas F. Y. Chen,
Muhammad Firmansyah Kasim,
Luke Ceurvorst,
Naren Ratan,
James Sadler,
Matthew C. Levy,
Raoul Trines,
Robert Bingham,
Peter Norreys
Abstract:
Proton radiography is a technique extensively used to resolve magnetic field structures in high energy density plasmas, revealing a whole variety of interesting phenomena such as magnetic reconnection and collisionless shocks found in astrophysical systems. Existing methods of analyzing proton radiographs give mostly qualitative results or specific quantitative parameters such as magnetic field st…
▽ More
Proton radiography is a technique extensively used to resolve magnetic field structures in high energy density plasmas, revealing a whole variety of interesting phenomena such as magnetic reconnection and collisionless shocks found in astrophysical systems. Existing methods of analyzing proton radiographs give mostly qualitative results or specific quantitative parameters such as magnetic field strength, and recent work showed that the line-integrated transverse magnetic field can be reconstructed in specific regimes where many simplifying assumptions were needed. Using artificial neural networks, we suggest a novel 3-D reconstruction method that works for a more general case. A proof of concept is presented here, with mean reconstruction errors of less than 5 percent even after introducing noise. We demonstrate that over the long term, this approach is more computationally efficient compared to other techniques. We also highlight the need for proton tomography because (i) certain field structures cannot be reconstructed from a single radiograph and (ii) errors can be further reduced when reconstruction is performed on radiographs generated by proton beams fired in different directions.
△ Less
Submitted 2 September, 2016; v1 submitted 19 August, 2016;
originally announced August 2016.
-
Quantitative shadowgraphy and proton radiography for large intensity modulations
Authors:
Muhammad Firmansyah Kasim,
Luke Ceurvorst,
Naren Ratan,
James Sadler,
Nicholas Chen,
Alexander Savert,
Raoul Trines,
Robert Bingham,
Philip N. Burrows,
Malte C. Kaluza,
Peter Norreys
Abstract:
Shadowgraphy is a technique widely used to diagnose objects or systems in various fields in physics and engineering. In shadowgraphy, an optical beam is deflected by the object and then the intensity modulation is captured on a screen placed some distance away. However, retrieving quantitative information from the shadowgrams themselves is a challenging task because of the non-linear nature of the…
▽ More
Shadowgraphy is a technique widely used to diagnose objects or systems in various fields in physics and engineering. In shadowgraphy, an optical beam is deflected by the object and then the intensity modulation is captured on a screen placed some distance away. However, retrieving quantitative information from the shadowgrams themselves is a challenging task because of the non-linear nature of the process. Here, a novel method to retrieve quantitative information from shadowgrams, based on computational geometry, is presented for the first time. This process can be applied to proton radiography for electric and magnetic field diagnosis in high-energy-density plasmas and has been benchmarked using a toroidal magnetic field as the object, among others. It is shown that the method can accurately retrieve quantitative parameters with error bars less than 10%, even when caustics are present. The method is also shown to be robust enough to process real experimental results with simple pre- and post-processing techniques. This adds a powerful new tool for research in various fields in engineering and physics for both techniques.
△ Less
Submitted 6 February, 2017; v1 submitted 14 July, 2016;
originally announced July 2016.
