arXiv subjects

Byoungchan Jang

Publications and source records attributed to Byoungchan Jang.

5 recordsLinked to original sources

arXiv2026

Bayesian optimization of stellarator alpha-particle confinement using data-informed parameter spaces and dimensionality reduction

Modern stellarators are typically designed by optimizing the shape of the plasma boundary surface, with the parameters taken to be Fourier amplitudes. Many promising optimization algorithms such as Bayesian methods require bound constraints on the parameters and are most efficient when each parameter is scaled similarly to the others. With the typical Fourier parameterization, it is unclear how to set these bounds: wide constraints lead to self-intersecting boundaries and frequent failures of the MHD equilibrium calculation, while tight bound constraints limit expressiveness. To address these issues, here we propose two new parameter spaces for stellarator optimization. Both begin with a dataset of existing stellarator boundaries. In the first approach, a quantile transformation is applied to each Fourier degree of freedom, mapping the data distribution to a uniform distribution on the unit interval. In the second approach, principal component analysis (PCA) is applied to points on the boundaries, followed by a quantile transformation. For both approaches, the transformed variables become the degrees of freedom, naturally bounded to [0, 1]. The PCA method has the additional benefit of dimensionality reduction, with high expressiveness for a small number of parameters. The methods are demonstrated via Bayesian optimization for good alpha-particle confinement with guiding-center tracing inside the optimization loop, using asynchronous parallelization. These optimizations yield stellarator configurations with excellent fast-particle confinement in fields that can be far from quasisymmetric or quasi-isodynamic.

physics.plasm-ph↗

arXiv2026

Deflation Techniques for Stellarator Equilibrium and Optimization

Stellarator optimization is a multi-objective, non-convex problem characterized by a complex objective landscape containing many local minima. The solution resulting from a single optimization is highly sensitive to factors such as the initial guess, objective weights, and the optimization method employed. However, merely varying these factors does not guarantee that a physically distinct minimum will be found; optimizations often fail to converge to good minima or simply return to the same or very similar local minima despite large-scale parameter scans. This paper presents a novel application of deflation methods to effectively explore this landscape. By modifying the objective function to penalize and "deflate" away already-found solutions, this technique encourages the optimizer towards attractive, distinct new minima while using a single initial guess and optimization setup. We provide a primer on deflation for nonlinear systems and non-convex optimization before applying it to non-axisymmetric equilibrium and stellarator optimization problems. Key results include the discovery of families of global equilibria with similar core characteristics and the convergence to helical core equilibria without prescient initial guesses. Furthermore, we demonstrate that augmenting stage-one stellarator and stage-two coil optimization with deflation constraints readily produces multiple high-quality, distinct solutions, establishing the method's efficacy and ease of use.

physics.plasm-ph↗

arXiv2025

Exponential Spectral Scaling: Robust and Efficient Stellarator Boundary Optimization via Mode-Dependent Scaling

Stellarator boundary optimization faces a fundamental numerical challenge: the extreme disparity between low- and high-mode amplitudes creates an optimization landscape in which direct full-spectrum approaches typically converge to poor local minima. Traditionally, this challenge has been addressed through a computationally expensive, multi-step Fourier continuation, in which low Fourier modes are optimized first, followed by the gradual incorporation of higher modes. We present Exponential Spectral Scaling (ESS), a technique that applies a mode-dependent exponential scaling factor to each Fourier mode. Our primary implementation uses the $L_{\infty}$ norm to determine the scaling pattern, creating a square spectral decay profile that effectively reduces the dynamic range of optimization variables from $10^{6}$--$10^{7}$ to $10^{2}$--$10^{3}$. This scaling aligns with the natural spectral decay of physically meaningful configurations and enables direct single-step optimization using the full spectrum of boundary Fourier modes. ESS eliminates arbitrary staging decisions and reduces computation time by a factor of $2$ to $5$ in benchmark cases. In addition to accelerating optimization, ESS improves robustness, reducing sensitivity to initial conditions and increasing confidence in avoiding local optima. We demonstrate the effectiveness of ESS across both quasi-axisymmetric (QA) and quasi-helically symmetric (QH) configurations, using two distinct optimization toolkits: SIMSOPT and DESC.

physics.plasm-ph↗

arXiv2023

Grad-Shafranov equilibria via data-free physics informed neural networks

A large number of magnetohydrodynamic (MHD) equilibrium calculations are often required for uncertainty quantification, optimization, and real-time diagnostic information, making MHD equilibrium codes vital to the field of plasma physics. In this paper, we explore a method for solving the Grad-Shafranov equation by using Physics-Informed Neural Networks (PINNs). For PINNs, we optimize neural networks by directly minimizing the residual of the PDE as a loss function. We show that PINNs can accurately and effectively solve the Grad-Shafranov equation with several different boundary conditions. We also explore the parameter space by varying the size of the model, the learning rate, and boundary conditions to map various trade-offs such as between reconstruction error and computational speed. Additionally, we introduce a parameterized PINN framework, expanding the input space to include variables such as pressure, aspect ratio, elongation, and triangularity in order to handle a broader range of plasma scenarios within a single network. Parametrized PINNs could be used in future work to solve inverse problems such as shape optimization.

physics.plasm-ph↗

arXiv2017

Prediction of a red nova outburst in KIC 9832227

We present the first identification of a candidate precursor for an imminent red nova. Our prediction is based on the example of the precursor to the red nova V1309 Sco, which was retrospectively found to be a contact binary with an exponentially decreasing period. We explore the use of this distinctive timing signature to identify precursors, developing the observational and analysis steps needed. We estimate that our Galaxy has roughly 1-10 observable precursors. Specifically, we lay out the observational case for KIC 9832227, which we identified as a tentative candidate two years ago (Molnar et al. 2015, AAS Meeting Abstracts 415.05). Orbital timing over the past two years has followed the tentative exponential fit. As of late 2015, the period time derivative went beyond the range found in other systems (dP/dt < 1x10^{-8}), a necessary criterion for a serious candidate. We estimate time of merger is the year 2022.2(7). Double absorption line spectra confirm directly the 0.458 d light curve period is a contact binary system and yield a mass ratio m_B/m_A = 0.228(3). Closer analysis of the Kepler timing data shows evidence of a component C with orbital period P_C = 590(8) days and m_C x sin i_C = 0.11 solar masses. An alternative interpretation of the long term timing trend, light travel time delay due to orbit around a distant component D, is ruled out by the spectroscopic data for any nondegenerate star. Additional measurements are needed to test further the merging hypothesis and to utilize fully this fortuitous opportunity.

astro-ph.SR↗