Search NASA⌕ Search

SEARCH · Search NASA

Results for “Nonlinear approximation”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 541 records · Page 30

Theory of molecule metal nano-particle interaction: Quantum description of plasmonic lasing

The recent quantum description of a few molecules interacting with plasmon excitations of a spherical metal nano-particle (MNP) as presented in the work of Zhang and May [Phys. Rev. B 89, 245441 (2014)] is extended to systems with up to 100 molecules. We demonstrate the possibility of multiple plasmon excitation and describe their conversion into far-field photons. The calculation of the steady-state photon emission spectrum results in an emission line-narrowing with an increasing number of molecules coupled to the MNP. This is considered as an essential criterion for the action of the molecule-MNP system as a nano-laser. To have exact results for systems with up to 20 molecules, we proceed as recently described by Richter et al. [Phys. Rev. B 91, 035306 (2015)] and study a highly symmetric system. It assumes an equatorial and regular position of identical molecules in such a way that their coupling is dominated by that to a single MNP dipole-plasmon excitation. Changing from the exact computation of the system’s complete density matrix to an approximate theory based on the reduced plasmon density matrix, systems with more than 100 molecules can be described. Finally, nonlinear rate equations are proposed which reproduce the mean number of excited plasmons in their dependence of the number of molecules and of the used pump rate. The second order intensity correlation function of emitted photons is related to the respective plasmon correlation function which approaches unity when the system starts lasing.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Compressed-sensing-based content-driven hierarchical reconstruction: Theory and application to C-arm cone-beam tomography

Purpose: This paper addresses the reconstruction of x-ray cone-beam computed tomography (CBCT) for interventional C-arm systems. Subsampling of CBCT is a significant issue with C-arms due to their slow rotation and to the low frame rate of their flat panel x-ray detectors. The aim of this work is to propose a novel method able to handle the subsampling artifacts generally observed with analytical reconstruction, through a content-driven hierarchical reconstruction based on compressed sensing. Methods: The central idea is to proceed with a hierarchical method where the most salient features (high intensities or gradients) are reconstructed first to reduce the artifacts these features induce. These artifacts are addressed first because their presence contaminates less salient features. Several hierarchical schemes aiming at streak artifacts reduction are introduced for C-arm CBCT: the empirical orthogonal matching pursuit approach with the ℓ{sub 0} pseudonorm for reconstructing sparse vessels; a convex variant using homotopy with the ℓ{sub 1}-norm constraint of compressed sensing, for reconstructing sparse vessels over a nonsparse background; homotopy with total variation (TV); and a novel empirical extension to nonlinear diffusion (NLD). Such principles are implemented with penalized iterative filtered backprojection algorithms. For soft-tissue imaging, the authors compare the use of TV and NLD filters as sparsity constraints, both optimized with the alternating direction method of multipliers, using a threshold for TV and a nonlinear weighting for NLD. Results: The authors show on simulated data that their approach provides fast convergence to good approximations of the solution of the TV-constrained minimization problem introduced by the compressed sensing theory. Using C-arm CBCT clinical data, the authors show that both TV and NLD can deliver improved image quality by reducing streaks. Conclusions: A flexible compressed-sensing-based algorithmic approach is proposed that is able to accommodate for a wide range of constraints. It is successfully applied to C-arm CBCT images that may not be so well approximated by piecewise constant functions.

60 APPLIED LIFE SCIENCES↗

On the Linearity of Local and Regional Temperature Changes from 1.5°C to 2°C of Global Warming

Because of the Paris Agreement it is imperative there is greater understanding of the consequences of limiting global warming to the target 1.5° and 2°C levels above preindustrial conditions. It is challenging to quantify changes across a small increment of global warming, so a pattern-scaling approach may be considered. Here we investigate the validity of such an approach by comprehensively examining how well local temperatures and warming trends in a 1.5°C world predict local temperatures at global warming of 2°C. Ensembles of transient coupled climate simulations from multiple models under different scenarios were compared and individual model responses were analyzed. For many places, the multimodel forced response of seasonal-average temperatures is approximately linear with global warming between 1.5° and 2°C. Yet, individual model results vary and large contributions from nonlinear changes in unforced variability or the forced response cannot be ruled out. In some regions, such as East Asia, models simulate substantially greater warming than is expected from linear scaling. Examining East Asia during boreal summer, we find that increased warming in the simulated 2°C world relative to scaling up from 1.5°C is related to reduced anthropogenic aerosol emissions. Our results indicate that, where forcings other than those due to greenhouse gas emissions change, the warming experienced in a 1.5°C world is a poor predictor for local climate at 2°C of global warming. In addition to the analysis of the linearity in the forced climate change signal, we find that natural variability remains a substantial contribution to uncertainty at these low-warming targets.

54 ENVIRONMENTAL SCIENCES↗

Constructing diabatic representations using adiabatic and approximate diabatic data – Coping with diabolical singularities

We have recently introduced a diabatization scheme, which simultaneously fits and diabatizes adiabatic ab initio electronic wave functions, Zhu and Yarkony J. Chem. Phys. 140, 024112 (2014). The algorithm uses derivative couplings in the defining equations for the diabatic Hamiltonian, H{sup d}, and fits all its matrix elements simultaneously to adiabatic state data. This procedure ultimately provides an accurate, quantifiably diabatic, representation of the adiabatic electronic structure data. However, optimizing the large number of nonlinear parameters in the basis functions and adjusting the number and kind of basis functions from which the fit is built, which provide the essential flexibility, has proved challenging. In this work, we introduce a procedure that combines adiabatic state and diabatic state data to efficiently optimize the nonlinear parameters and basis function expansion. Further, we consider using direct properties based diabatizations to initialize the fitting procedure. To address this issue, we introduce a systematic method for eliminating the debilitating (diabolical) singularities in the defining equations of properties based diabatizations. We exploit the observation that if approximate diabatic data are available, the commonly used approach of fitting each matrix element of H{sup d} individually provides a starting point (seed) from which convergence of the full H{sup d} construction algorithm is rapid. The optimization of nonlinear parameters and basis functions and the elimination of debilitating singularities are, respectively, illustrated using the 1,2,3,4{sup 1}A states of phenol and the 1,2{sup 1}A states of NH{sub 3}, states which are coupled by conical intersections.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Thermodynamics and luminosities of rainbow black holes

Doubly special relativity (DSR) is an effective model for encoding quantum gravity in flat spacetime. As result of the nonlinearity of the Lorentz transformation, the energy-momentum dispersion relation is modified. One simple way to import DSR to curved spacetime is ''Gravity's rainbow'', where the spacetime background felt by a test particle would depend on its energy. Focusing on the ''Amelino-Camelia dispersion relation'' which is E{sup 2} = m{sup 2}+p{sup 2}[1−η(E/m{sub p}){sup n}] with n > 0, we investigate the thermodynamical properties of a Schwarzschild black hole and a static uncharged black string for all possible values of η and n in the framework of rainbow gravity. It shows that there are non-vanishing minimum masses for these two black holes in the cases with η < 0 and n ≥ 2. Considering effects of rainbow gravity on both the Hawking temperature and radius of the event horizon, we use the geometric optics approximation to compute luminosities of a 2D black hole, a Schwarzschild one and a static uncharged black string. It is found that the luminosities can be significantly suppressed or boosted depending on the values of η and n.

79 ASTRONOMY AND ASTROPHYSICS↗

Cosmological perturbation theory in 1+1 dimensions

Many recent studies have highlighted certain failures of the standard Eulerian-space cosmological perturbation theory (SPT). Its problems include (1) not capturing large-scale bulk flows [leading to an O( 1) error in the 1-loop SPT prediction for the baryon acoustic peak in the correlation function], (2) assuming that the Universe behaves as a pressureless, inviscid fluid, and (3) treating fluctuations on scales that are non-perturbative as if they were. Recent studies have highlighted the successes of perturbation theory in Lagrangian space or theories that solve equations for the effective dynamics of smoothed fields. Both approaches mitigate some or all of the aforementioned issues with SPT. We discuss these physical developments by specializing to the simplified 1D case of gravitationally interacting sheets, which allows us to substantially reduces the analytic overhead and still (as we show) maintain many of the same behaviors as in 3D. In 1D, linear-order Lagrangian perturbation theory ('the Zeldovich approximation') is exact up to shell crossing, and we prove that n{sup th}-order Eulerian perturbation theory converges to the Zeldovich approximation as narrow ∞. In no 1D cosmology that we consider (including a CDM-like case and power-law models) do these theories describe accurately the matter power spectrum on any mildly nonlinear scale. We find that theories based on effective equations are much more successful at describing the dynamics. Finally, we discuss many topics that have recently appeared in the perturbation theory literature such as beat coupling, the shift and smearing of the baryon acoustic oscillation feature, and the advantages of Fourier versus configuration space. Our simplified 1D case serves as an intuitive review of these perturbation theory results.

79 ASTRONOMY AND ASTROPHYSICS↗

Turbulent Regimes in Collisions of 3D Alfvén-wave Packets

Using three-dimensional gyrofluid simulations, we revisit the problem of Alfvén-wave (AW) collisions as building blocks of the Alfvénic turbulent cascade and their interplay with magnetic reconnection at magnetohydrodynamic (MHD) scales. Depending on the large-scale value of the nonlinearity parameter χ 0 (the ratio between the AW linear propagation time and nonlinear turnover time), different regimes are observed. For strong nonlinearities (χ 0 ~ 1), turbulence is consistent with a dynamically aligned, critically balanced cascade—fluctuations exhibit a scale-dependent alignment $\sin {\theta }_{{k}_{\perp }}\propto {k}_{\perp }^{-1/4}$, resulting in a ${k}_{\perp }^{-3/2}$ spectrum and ${k}_{\parallel }\propto {k}_{\perp }^{1/2}$ spectral anisotropy. At weaker nonlinearities (small χ 0 ), a spectral break marking the transition between a large-scale weak regime and a small-scale ${k}_{\perp }^{-11/5}$ tearing-mediated range emerges, implying that dynamic alignment occurs also for weak nonlinearities. At χ 0 < 1 the alignment angle ${\theta }_{{k}_{\perp }}$ shows a stronger scale dependence than in the χ 0 ~ 1 regime, namely $\sin {\theta }_{{k}_{\perp }}\propto {k}_{\perp }^{-1/2}$ at χ 0 ~ 0.5, and $\sin {\theta }_{{k}_{\perp }}\propto {k}_{\perp }^{-1}$ at χ 0 ~ 0.1. Dynamic alignment in the weak regime also modifies the large-scale spectrum, scaling approximately as ${k}_{\perp }^{-3/2}$ for χ 0 ~ 0.5 and as ${k}_{\perp }^{-1}$ for χ 0 ~ 0.1. A phenomenological theory of dynamically aligned turbulence at weak nonlinearities that can explain these spectra and the transition to the tearing-mediated regime is provided; at small χ 0 , the strong scale dependence of the alignment angle combines with the increased lifetime of turbulent eddies to allow tearing to onset and mediate the cascade at scales that can be larger than those predicted for a critically balanced cascade by several orders of magnitude. Such a transition to tearing-mediated turbulence may even supplant the usual weak-to-strong transition.

79 ASTRONOMY AND ASTROPHYSICS↗

The 2013 February 17 Sunquake in the Context of the Active Region's Magnetic Field Configuration

Sunquakes are created by the hydrodynamic response of the lower atmosphere to a sudden deposition of energy and momentum. In this study, we investigate a sunquake that occurred in NOAA active region 11675 on 2013 February 17. Observations of the corona, chromosphere, and photosphere are brought together for the first time with a nonlinear force-free model of the active region’s magnetic field in order to probe the magnetic environment in which the sunquake was initiated. We find that the sunquake was associated with the destabilization of a flux rope and an associated M-class GOES flare. Active region 11675 was in its emergence phase at the time of the sunquake and photospheric motions caused by the emergence heavily modified the flux rope and its associated quasi-separatrix layers, eventually triggering the flux rope’s instability. The flux rope was surrounded by an extended envelope of field lines rooted in a small area at the approximate position of the sunquake. We argue that the configuration of the envelope, by interacting with the expanding flux rope, created a “magnetic lens” that may have focussed energy on one particular location of the photosphere, creating the necessary conditions for the initiation of the sunquake.

79 ASTRONOMY AND ASTROPHYSICS↗

Three dimensional full-wave nonlinear acoustic simulations: Applications to ultrasound imaging

Characterization of acoustic waves that propagate nonlinearly in an inhomogeneous medium has significant applications to diagnostic and therapeutic ultrasound. The generation of an ultrasound image of human tissue is based on the complex physics of acoustic wave propagation: diffraction, reflection, scattering, frequency dependent attenuation, and nonlinearity. The nonlinearity of wave propagation is used to the advantage of diagnostic scanners that use the harmonic components of the ultrasonic signal to improve the resolution and penetration of clinical scanners. One approach to simulating ultrasound images is to make approximations that can reduce the physics to systems that have a low computational cost. Here a maximalist approach is taken and the full three dimensional wave physics is simulated with finite differences. This paper demonstrates how finite difference simulations for the nonlinear acoustic wave equation can be used to generate physically realistic two and three dimensional ultrasound images anywhere in the body. A specific intercostal liver imaging scenario for two cases: with the ribs in place, and with the ribs removed. This configuration provides an imaging scenario that cannot be performed in vivo but that can test the influence of the ribs on image quality. Several imaging properties are studied, in particular the beamplots, the spatial coherence at the transducer surface, the distributed phase aberration, and the lesion detectability for imaging at the fundamental and harmonic frequencies. The results indicate, counterintuitively, that at the fundamental frequency the beamplot improves due to the apodization effect of the ribs but at the same time there is more degradation from reverberation clutter. At the harmonic frequency there is significantly less improvement in the beamplot and also significantly less degradation from reverberation. It is shown that even though simulating the full propagation physics is computationally challenging it is necessary to quantify ultrasound image quality and its sources of degradation.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Optimization of nonlinear turbulence in stellarators

We present new stellarator equilibria that have been optimized for reduced turbulent transport using nonlinear gyrokinetic simulations within the optimization loop. The optimization routine involves coupling the pseudo-spectral GPU-native gyrokinetic code GX with the stellarator equilibrium and optimization code DESC . Since using GX allows for fast nonlinear simulations, we directly optimize for reduced nonlinear heat fluxes. To handle the noisy heat flux traces returned by these simulations, we employ the simultaneous perturbation stochastic approximation (SPSA) method that only uses two objective function evaluations for a simple estimate of the gradient. We show several examples that optimize for both reduced heat fluxes and good quasi-symmetry as a proxy for low neoclassical transport. Finally, we run full transport simulations using the T3D stellarator transport code to evaluate the changes in the macroscopic profiles.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Facile and quantitative estimation of strain in nanobubbles with arbitrary symmetry in 2D semiconductors verified using hyperspectral nano-optical imaging

When layers of van der Waals materials are deposited via exfoliation or viscoelastic stamping, nanobubbles are sometimes created from aggregated trapped fluids. Though they can be considered a nuisance, nanobubbles have attracted scientific interest in their own right owing to their ability to generate large in-plane strain gradients that lead to rich optoelectronic phenomena, especially in the semiconducting transition metal dichalcogenides. Determination of the strain within the nanobubbles, which is crucial to understanding these effects, can be approximated using elasticity theory. However, the Föppl–von Kármán equations that describe strain in a distorted thin plate are highly nonlinear and often necessitate assuming circular symmetry to achieve an analytical solution. Here, we present an easily implemented numerical method to solve for strain tensors of nanobubbles with arbitrary symmetry in 2D crystals. The method only requires topographic information from atomic force microscopy and the Poisson ratio of the 2D material. We verify that this method reproduces the strain for circularly symmetric nanobubbles that have known analytical solutions. Finally, we use the method to reproduce the Grüneisen parameter of the E' mode for 1L-WS2 nanobubbles on template-stripped Au by comparing the derived strain with measured Raman shifts from tip-enhanced Raman spectroscopy, demonstrating the utility of our method for estimating localized strain in 2D crystals.

2D materials↗

Mitigation of stimulated Raman scattering in the kinetic regime by external magnetic fields

Here, we show via particle-in-cell simulations that small normalized magnetic fields (ω c /ω p << 1) can significantly modify the evolution of backward stimulated Raman scattering (SRS) in the kinetic regime due to the enhanced dissipation of nonlinear electron plasma waves propagating perpendicular to magnetic fields. A magnetic field applied perpendicularly to the electron plasma wave (and driving light wave) increases the SRS threshold for kinetic inflation and decreases the amount of reflectivity when SRS is driven significantly above threshold. Analysis indicates that this arises because trapped electrons are accelerated as they surf across the wave, leading to the continual dissipation of the electron plasma waves over a wider range of wave amplitudes. The reduction in SRS reflectivity is most significant for a purely perpendicular field, although reduction also occurs for other angles; a parallel field can slightly increase single-speckle SRS but decreases multispeckle SRS. These simulations demonstrate the significance of magnetic-field contributions to nonlinear electron plasma wave damping with respect to nonlinear parametric decay instabilities; the simulation parameters are directly relevant for SRS in inertial confinement fusion devices and indicate that approximately 30 tesla magnetic fields might significantly reduce SRS backscatter.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Data-Driven Reduced-Order Modeling of Convective Heat Transfer in Porous Media

This work presents a data-driven Reduced-Order Model (ROM) for parametric convective heat transfer problems in porous media. The intrusive Proper Orthogonal Decomposition aided Reduced-Basis (POD-RB) technique is employed to reduce the porous medium formulation of the incompressible Reynolds-Averaged Navier–Stokes (RANS) equations coupled with heat transfer. Instead of resolving the exact flow configuration with high fidelity, the porous medium formulation solves a homogenized flow in which the fluid-structure interactions are captured via volumetric flow resistances with nonlinear, semi-empirical friction correlations. A supremizer approach is implemented for the stabilization of the reduced fluid dynamics equations. The reduced nonlinear flow resistances are treated using the Discrete Empirical Interpolation Method (DEIM), while the turbulent eddy viscosity and diffusivity are approximated by adopting a Radial Basis Function (RBF) interpolation-based approach. The proposed method is tested using a 2D numerical model of the Molten Salt Fast Reactor (MSFR), which involves the simulation of both clean and porous medium regions in the same domain. For the steady-state example, five model parameters are considered to be uncertain: the magnitude of the pumping force, the external coolant temperature, the heat transfer coefficient, the thermal expansion coefficient, and the Prandtl number. For transient scenarios, on the other hand, the coastdown-time of the pump is the only uncertain parameter. The results indicate that the POD-RB-ROMs are suitable for the reduction of similar problems. The relative L2 errors are below 3.34% for every field of interest for all cases analyzed, while the speedup factors vary between 54 (transient) and 40,000 (steady-state).

German, Péter (ORCID:0000000307285283)↗

SUNSPOT AND STARSPOT LIFETIMES IN A TURBULENT EROSION MODEL

Quantitative models of sunspot and starspot decay predict the timescale of magnetic diffusion and may yield important constraints in stellar dynamo models. Motivated by recent measurements of starspot lifetimes, we investigate the disintegration of a magnetic flux tube by nonlinear diffusion. Previous theoretical studies are extended by considering two physically motivated functional forms for the nonlinear diffusion coefficient D : an inverse power-law dependence D ∝ B {sup −ν} and a step-function dependence of D on the magnetic field magnitude B . Analytical self-similar solutions are presented for the power-law case, including solutions exhibiting “super fast” diffusion. For the step-function case, the heat-balance integral method yields approximate solutions, valid for moderately suppressed diffusion in the spot. The accuracy of the resulting solutions is confirmed numerically, using a method which provides an accurate description of long-time evolution by imposing boundary conditions at infinite distance from the spot. The new models may allow insight into the differences and similarities between sunspots and starspots.

79 ASTRONOMY AND ASTROPHYSICS↗

Development of Segregated Thermal-Hydraulics Solvers in MOOSE

The simulation of fluid flows is an essential part of the design and analysis of nuclear systems. Algorithms able to simulate flows at different fidelity levels are available in the Multiphysics Object-Oriented Simulation Environment (MOOSE) and MOOSE-based applications such as Pronghorn \cite{novak2018pronghorn}, Pronghorn-Subchannel, RELAP-7, and SAM. Currently, significant effort is being invested in the development of coarse-mesh Computational Fluid Dynamics (CFD) capabilities within MOOSE and Pronghorn for the simulation of Generation IV nuclear reactors. Traditionally, the solution algorithms in MOOSE have relied on Newton or quasi-Newton methods (such as the preconditioned Jacobian-free Newton-Krylov method) where residuals and Jacobians (or approximations thereof) are constructed. Both Newton and quasi-Newton methods require the solution of a linear system at each nonlinear Newton iteration with the Jacobian as the system matrix. The Jacobian contains blocks originating from all variables in the problem (i.e., for thermal-hydraulics at least pressure, velocities, and temperature). Due to the formulation of the problem in a general multiphysics setting on unstructured mesh, creating a good preconditioner for the linear system can be challenging, thus many fluid applications have utilized direct solver-based methods such as LU factorization. However, with increasing system size and complexity in multi-dimensional problems, the direct solution of linear systems becomes computationally expensive both in execution time and and memory. For this reason, recent effort has focused on adapting segregated solution algorithms for CFD problems in MOOSE. These algorithms use fixed-point iteration between segregated systems whose assembly and preconditioning are easier those of the monolithic system. Initial results show that the segregated solution algorithm outperforms the monolithic approach in terms of memory usage and for large 3D problems in terms of CPU time as well.

42 ENGINEERING↗

Development of Segregated Thermal-Hydraulics Solvers in MOOSE

The simulation of fluid flows is an essential part of the design and analysis of nuclear systems. Algorithms able to simulate flows at different fidelity levels are available in the Multiphysics Object-Oriented Simulation Environment (MOOSE) and MOOSE-based applications such as Pronghorn \cite{novak2018pronghorn}, Pronghorn-Subchannel, RELAP-7, and SAM. Currently, significant effort is being invested in the development of coarse-mesh Computational Fluid Dynamics (CFD) capabilities within MOOSE and Pronghorn for the simulation of Generation IV nuclear reactors. Traditionally, the solution algorithms in MOOSE have relied on Newton or quasi-Newton methods (such as the preconditioned Jacobian-free Newton-Krylov method) where residuals and Jacobians (or approximations thereof) are constructed. Both Newton and quasi-Newton methods require the solution of a linear system at each nonlinear Newton iteration with the Jacobian as the system matrix. The Jacobian contains blocks originating from all variables in the problem (i.e., for thermal-hydraulics at least pressure, velocities, and temperature). Due to the formulation of the problem in a general multiphysics setting on unstructured mesh, creating a good preconditioner for the linear system can be challenging, thus many fluid applications have utilized direct solver-based methods such as LU factorization. However, with increasing system size and complexity in multi-dimensional problems, the direct solution of linear systems becomes computationally expensive both in execution time and and memory. For this reason, recent effort has focused on adapting segregated solution algorithms for CFD problems in MOOSE. These algorithms use fixed-point iteration between segregated systems whose assembly and preconditioning are easier those of the monolithic system. Initial results show that the segregated solution algorithm outperforms the monolithic approach in terms of memory usage and for large 3D problems in terms of CPU time as well.

42 ENGINEERING↗

Wavenumber shift due to nonlinear plasma and wave interaction

Wavenumber shift of the ion Bernstein wave has been observed in the particle-in-cell simulations when the input power of the injected wave is sufficiently large. It is demonstrated that the increase of the total kinetic energy of ions, including both the thermal energy related to the random thermal motion and the oscillation energy due to the coherent motion with the wave, gives rise to such change of the wavenumber. However, the velocity distribution function of the ions can approximately be fitted as a Maxwellian distribution function, and thus, the linear dispersion relation still holds, provided that the initial ion temperature is replaced by the effective temperature measured in the simulation.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Machine Learning-Based Model Predictive Control of Two-Time-Scale Systems

In this study, we present a general form of nonlinear two-time-scale systems, where singular perturbation analysis is used to separate the dynamics of the slow and fast subsystems. Machine learning techniques are utilized to approximate the dynamics of both subsystems. Specifically, a recurrent neural network (RNN) and a feedforward neural network (FNN) are used to predict the slow and fast state vectors, respectively. Moreover, we investigate the generalization error bounds for these machine learning models approximating the dynamics of two-time-scale systems. Next, under the assumption that the fast states are asymptotically stable, our focus shifts toward designing a Lyapunov-based model predictive control (LMPC) scheme that exclusively employs the RNN to predict the dynamics of the slow states. Additionally, we derive sufficient conditions to guarantee the closed-loop stability of the system under the sample-and-hold implementation of the controller. A nonlinear chemical process example is used to demonstrate the theory. In particular, two RNN models are constructed: one to model the full two-time-scale system and the other to predict solely the slow state vector. Both models are integrated within the LMPC scheme, and we compare their closed-loop performance while assessing the computational time required to execute the LMPC optimization problem.

97 MATHEMATICS AND COMPUTING↗