High-fidelity finite-element simulations can provide accurate numerical predictions for side-branch resonators, but large simulation datasets are expensive to generate and purely data-driven surrogates may become unreliable when simulation-labelled data are scarce. This study develops an analytical-prior learning framework that reuses a low-cost analytical model to improve data efficiency under limited high-fidelity simulation budgets. Two complementary routes are considered. When the analytical model remains available at inference, it is retained as an explicit baseline and the simulation data are used to learn only the analytical-to-simulation discrepancy. When a self-contained predictor is required, the analytical mapping is first distilled from abundant low-cost evaluations into a learned prior and then calibrated with the limited simulation data. The framework is evaluated on rectangular side-branch Helmholtz resonators using 86 simulation-labelled geometries and 8,998 non-overlapping analytical-only geometries. The analytical model achieved a mean absolute error (MAE) of 1.333 Hz. Direct support vector regression (SVR) achieved 3.375 Hz, while residual SVR reduced the MAE to 0.426 Hz. A direct multilayer perceptron (MLP) achieved 1.109 Hz, whereas analytical-prior pretraining reduced the error to 0.556 Hz with frozen-prior residual adaptation and 0.371 Hz with full-model fine-tuning. Across training budgets of 20 to 70 simulation-labelled cases, both analytical correction and analytical-prior pretraining consistently improved data efficiency relative to direct learning. These results show that analytical prior information can substantially improve high-fidelity prediction when simulation data are scarce, with explicit correction and prior distillation serving complementary deployment needs.
The inverse design of acoustic metamaterials with targeted resonance characteristics remains a significant challenge due to the high-dimensional and nonlinear design space as well as strict physical constraints such as subwavelength operation and viscothermal losses. A hybrid neural-physics
inverse design framework is presented, integrating a conditional variational autoencoder (cVAE) with a physics-based analytical model to generate and refine metamaterial unit cells. Artificial target sound absorption coefficient (SAC) spectra are used to train the cVAE to learn a compact latent
representation, enabling fast and differentiable generation of candidate geometries. Latent-space optimization minimizes the mismatch between predicted and target SAC, yielding neural-optimal designs. To ensure physical consistency, these candidates are further refined using a covariance matrix
adaptation evolution strategy (CMA-ES) coupled with a Numba-accelerated analytical model of Helmholtz resonators with rough necks, which accounts for viscothermal losses and end corrections. This hybrid approach enables precise control of resonance frequency, peak absorption, and bandwidth.
Benchmark studies show improved SAC matching, faster convergence, and greater robustness compared with purely data-driven or purely analytical approaches. The proposed framework provides an efficient tool for the inverse design of subwavelength acoustic metamaterials with potential applications
in aircraft cabin noise mitigation, building acoustics, and advanced sound control systems.
T. C. Kone, S. Ghinet, A. Grewal· INTER-NOISE and NOISE-CON Co...· 0 citations
Repeated prediction of acoustic fields from spatially distributed boundary excitation is computationally expensive when each source realization requires a new wave simulation. This work introduces a quadrature-aware complex-linear boundary operator (CLBO) that maps complex normal velocity on a vibrating surface to complex pressure at receiver locations. The model couples learned source and receiver basis functions through an explicit complex surface-quadrature contraction, so the boundary excitation enters linearly by construction. This preserves complex superposition, homogeneity, and zero response to zero excitation, while representing the source through coordinates, normals, and quadrature weights rather than a fixed flattened input vector. Reference data were generated using a verified three-dimensional multiple-relaxation-time (MRT) lattice Boltzmann solver and stored in a solver-agnostic boundary-to-field format. CLBO was compared with a fixed-sensor complex DeepONet under matched case splits and optimization settings, with additional tests of structural consistency, receiver-coordinate interpolation, source discretization, source-family holdout, label efficiency, physics-informed ablations, unseen source mixtures, and computational cost. Across five training seeds, CLBO achieved a mean complex relative field error of 0.184 +/- 0.00771, compared with 0.367 +/- 0.00742 for DeepONet. Its measured source-superposition error was 1.31 x 10^-7, and its mean error on newly simulated mixed-source cases was 0.237, compared with 0.415 for DeepONet. Inference was 1.83 x 10^4 faster than the reference calculation for the reported query size. These results show that enforcing the known complex-linear boundary-to-field structure improves physical consistency and generalization under distributed acoustic excitation.
Conventional acoustic impedance inversion methods have long faced technical bottlenecks such as inaccurate wavelet estimation and strong dependence on initial models. Although existing deep learning approaches can partially alleviate these problems, they often compromise model simplicity and training efficiency, while introducing new challenges such as limited generalizability and heavy reliance on labeled data. To overcome these limitations, this study proposes a lightweight inversion framework that tightly integrates physics-driven and data-driven paradigms. The physics-driven component adopts a neural network architecture largely consistent with traditional modeling processes, enabling direct optimization of physically meaningful parameters through backpropagation, thereby avoiding the construction of excessively complex inverse operators. Meanwhile, regularization methods are introduced to enforce geological prior knowledge (that is, the “layered geological model” assumption) on the network parameters, improving the spatial continuity of the reconstructed impedance models. The data-driven component employs an enhanced 2D U-Net integrated with Class Activation Mapping (U-Net-CAM) to generate accurate reference models from sparse well-log data. Tests on both synthetic and field datasets demonstrate the advantages of the proposed method: (1) physically interpretable network design; (2) strong robustness to noise and reduced dependence on training data; (3) higher accuracy and better spatial continuity compared to conventional and purely data-driven methods. This work provides a new perspective for addressing long-standing challenges in seismic impedance inversion.
Abstract. High-fidelity computational aeroacoustic simulations are often too computationally expensive for rapid evaluation of aircraft noise, especially when assessing full-system acoustic footprints during operations. This work introduces a physics-informed multipolar-expansion methodology that generates equivalent acoustic sources capable of accurately reproducing complex pressure fields in both near- and far-field regions at a fraction of the computational cost. The method employs expansions from monopoles to higher-order terms for noise modeling and prediction, using spherical harmonics as a complete basis of the Helmholtz equation. Acoustic moments coefficients are identified through linear regression analysis of pressure data obtained from numerical or experimental sampling, enabling the accurate reconstruction of the acoustic field once the coefficients are determined. The methodology is assessed using two representative aeroacoustic problems: a helicopter rotor operating in Blade–Vortex Interaction conditions and a subsonic jet at Mach 0.9 with varying nozzle temperatures. The results show excellent agreement with reference data, maintaining amplitude accuracy up to the far field while drastically reducing computational time. The proposed reduced-order model demonstrates strong robustness and generalizability, offering a fast and physically grounded tool for aerodynamic optimization, real-time control, and digital-twin frameworks, enabling efficient and scalable noise prediction in complex aeroacoustic environments.
Giorgio Palma· Materials Research Proceedin...· 0 citations
Neural surrogates offer a promising route to accelerating computationally expensive simulations governed by partial differential equations across science and industry. Their practical deployment, however, is limited by unreliable predictions under out-of-distribution (OOD) conditions. We develop a solver-coupled surrogate-Newton framework that uses surrogate predictions as high-quality initial guesses for Newton-Krylov iterations, thereby combining rapid global flow-field prediction with high-accuracy numerical convergence at the terminal stage. On an OOD benchmark comprising geometries sampled from actual transonic airfoil optimization trajectories, the framework lowers the median residual L_2 ratio by over seven orders of magnitude while substantially reducing field and aerodynamic errors. In practical supercritical airfoil optimization, it improves online prediction reliability while achieving a 15.5-fold generation-level speedup over CFD. We further test the framework's extension to three dimensions using a flying-wing dataset. Together, these studies demonstrate the potential of surrogate-Newton coupling to deliver accurate, efficient and scalable steady CFD across industrial workflows.
Ming Lei, Weishao Tang, Yufei Zhang et al.· 0 citations
This paper presents a data-efficient surrogate modeling framework for electromagnetic structures and components based on a compressed formulation of vector-valued kernel ridge regression (VV-KRR). The method is intended for small-data regimes, where training samples generated through full-wave or circuit simulations are computationally expensive and therefore limited in number. By combining a compressed spectral formulation with a fully data-driven output-kernel construction, the proposed approach enables accurate multi-output modeling with reduced memory footprint, lower training cost, and improved computational scalability. The framework retains the ability of VV-KRR to capture cross-output correlations and to model high-dimensional responses, while featuring a simple training procedure and good numerical robustness. It is particularly attractive for simulation-driven design workflows in high-speed interconnect, RF, and microwave applications, where data generation is often the dominant computational cost. The method is validated on three representative electromagnetic modeling problems, demonstrating that accurate and robust surrogate models can be obtained even in data-constrained scenarios.
N. Soleimani, I. Stievano, Riccardo Trinchero· IEEE Access· 0 citations
A new machine-learning framework aims to improve the success rate of computational protein design while moving away from results that reproduce sequences found in nature.
MIT News · Artificial Intelligence· news.mit.eduAug 24, 2026
