A pole-residue polynomial-chaos neural network for stochastic frequency response of MDOF systems
Abstract
The statistics of the frequency response of an uncertain structure are most demanding to compute near lightly damped resonances. As the parameters vary, the resonant frequencies migrate, so at a fixed frequency the response depends sharply on the parameters, and a polynomial surrogate converges slowly there. We propose to work on its partial-fraction representation characterized by poles and residues. The surrogate is an encoder–decoder with a physical bottleneck, in which an encoder maps the random parameters into the spectral data and a fixed analytic pole–residue layer decodes them into the response. We introduce two encoders of the spectral data, a direct polynomial-chaos map, PR-PCE, and a neural-network encoder that takes the chaos basis as its feature map, PR-PCNN. We develop the construction for a random vector indexed by the frequency. A complementary route learns the same spectral code from sampled responses alone, independently of the system matrices. The framework is verified on two systems, a three-degree-of-freedom structure and a clamped–free beam whose flexural rigidity is a random field. On both, the linear PR-PCE map is the most accurate, the neural encoders follow it, and a frequency-wise chaos expansion trails them. The representation is also benchmarked against a stochastic frequency transformation and a per-realisation vector fit. For these smooth spectral data the linear map stays the more accurate, and the value of the neural encoder lies in its flexibility, since it also learns the code end-to-end and from sampled responses alone.
Related
- Stochastic Analysis of Laser Phase Noise in Photonics-Assisted Radio Frequency Beamforming
- Black Holes as Frequency-Dependent Filters of Stochastic Gravitational Waves
- Prediction of high-frequency thermo-acoustic instabilities with a convective flame response model
- Recovery of event boundaries and Brillouin frequency shift in STFT-BOTDR using variable projection with spatial response constraint
- Stochastic evolutionary dynamics in structured populations: Network degree governs evolutionary stability
BibTeX
@article{adhikari2026poleresidue,
title = {{A pole-residue polynomial-chaos neural network for stochastic frequency response of MDOF systems}},
author = {Sondipon Adhikari and Eric Jacquelin},
journal = {Computational Mechanics},
year = {2026},
month = nov,
doi = {10.1007/s00466-026-02877-5},
url = {https://hal.science/hal-05851289v1},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - A pole-residue polynomial-chaos neural network for stochastic frequency response of MDOF systems
AU - Sondipon Adhikari
AU - Eric Jacquelin
PY - 2026
DA - 2026/11/02
JO - Computational Mechanics
DO - 10.1007/s00466-026-02877-5
UR - https://hal.science/hal-05851289v1
AB - The statistics of the frequency response of an uncertain structure are most demanding to compute near lightly damped resonances. As the parameters vary, the resonant frequencies migrate, so at a fixed frequency the response depends sharply on the parameters, and a polynomial surrogate converges slowly there. We propose to work on its partial-fraction representation characterized by poles and residues. The surrogate is an encoder–decoder with a physical bottleneck, in which an encoder maps the random parameters into the spectral data and a fixed analytic pole–residue layer decodes them into the response. We introduce two encoders of the spectral data, a direct polynomial-chaos map, PR-PCE, and a neural-network encoder that takes the chaos basis as its feature map, PR-PCNN. We develop the construction for a random vector indexed by the frequency. A complementary route learns the same spectral code from sampled responses alone, independently of the system matrices. The framework is verified on two systems, a three-degree-of-freedom structure and a clamped–free beam whose flexural rigidity is a random field. On both, the linear PR-PCE map is the most accurate, the neural encoders follow it, and a frequency-wise chaos expansion trails them. The representation is also benchmarked against a stochastic frequency transformation and a per-realisation vector fit. For these smooth spectral data the linear map stays the more accurate, and the value of the neural encoder lies in its flexibility, since it also learns the code end-to-end and from sampled responses alone.
SN - 0178-7675
SN - 1432-0924
ER -