Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “reduced order systems”

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 55 records · Page 3

Segmented multigrid domain decomposition solutions for three dimensional viscous recirculating flows

A segmented multigrid domain decomposition strategy is combined with a pressure-based form of flux-vector discretization for 3D incompressible and compressible viscous flow applications. A pressure-based form of flux-vector splitting is applied to the Navier-Stokes (NS) equations, which are represented by an implicit lowest-order reduced NS system and a purely diffusive higher-order deferred corrector. A trapezoidal or boxlike form of discretization insures that all mass conservation properties are satisfied at interfacial and outflow boundaries, even for this primitive-variable nonstaggered grid computation. Improvements in gridding strategy are presented by allowing for disjoint subdomains that provide optimal resolution of disparate flow features.

Srinivasan, Kumar↗

Bayesian reduced-order deep learning surrogate model for dynamic systems described by partial differential equations

We propose a reduced-order deep-learning surrogate model for dynamic systems described by time-dependent partial differential equations. This method employs space–time Karhunen–Loève expansions (KLEs) of the state variables and space-dependent KLEs of space-varying parameters to identify the reduced (latent) dimensions. Subsequently, a deep neural network (DNN) is used to map the parameter latent space to the state variable latent space. An approximate Bayesian method is developed for uncertainty quantification (UQ) in the proposed KL-DNN surrogate model. The KL-DNN method is tested for the linear advection–diffusion and nonlinear diffusion equations, and the Bayesian approach for UQ is compared with the deep ensembling (DE) approach, commonly used for quantifying uncertainty in DNN models. It was found that the approximate Bayesian method provides a more informative distribution of the PDE solutions in terms of the coverage of the reference PDE solutions (the percentage of nodes where the reference solution is within the confidence interval predicted by the UQ methods) and log predictive probability. The DE method is found to underestimate uncertainty and introduce bias. For the nonlinear diffusion equation, we compare the KL-DNN method with the Fourier Neural Operator (FNO) method and find that KL-DNN is 10% more accurate and needs less training time than the FNO method.

97 MATHEMATICS AND COMPUTING↗

Reduced-Order Modeling and Parameter Identification of Wind Tunnel Measurement Systems

We present a method to develop a physics-based, reduced-order model of a wind tunnel measurement system (including a sting, strain gage force balance, and test article) that can be used to predict the dynamics of the system. This reduced-order model is combined with a simple finite element beam model of a sting to estimate the dynamics of the full assembly. We make comparisons between a full finite element model and the hybrid reduced-order model to show that this hybrid reduced-order model is capable of predicting the first six natural frequencies to within 10% error. This technique could be used to identify reduced-order parameters for a large number of balances and stings, which could then be used to estimate the dynamics of different measurement assemblies.

Reduced-order Modeling↗

Reduced-Order Modeling and Parameter Identification of Wind Tunnel Measurement Systems

We present a method to develop a physics-based, reduced-order model of a wind tunnel measurement system (including a sting, strain gage force balance, and test article) that can be used to predict the dynamics of the system. This reduced-order model is combined with a simple finite element beam model of a sting to estimate the dynamics of the full assembly. We make comparisons between a full finite element model and the hybrid reduced-order model to show that this hybrid reduced-order model is capable of predicting the first six natural frequencies to within 10% error. This technique could be used to identify reduced-order parameters for a large number of balances and stings, which could then be used to estimate the dynamics of different measurement assemblies.

Reduced-order Modeling↗

Data-driven reduced-order models for port-Hamiltonian systems with operator inference

Hamiltonian operator inference has been developed in Sharma et al. (2022) to learn structure-preserving reduced-order models (ROMs) for Hamiltonian systems. The method constructs a low-dimensional model using only data and knowledge of the functional form of the Hamiltonian. The resulting ROMs preserve the intrinsic structure of the system, ensuring that the mechanical and physical properties of the system are maintained. In this work, we extend this approach to port-Hamiltonian systems, which generalize Hamiltonian systems by including energy dissipation, external input, and output. Based on snapshots of the system’s state and output, together with the information about the functional form of the Hamiltonian, reduced operators are inferred through optimization and are then used to construct data-driven ROMs. To further alleviate the complexity of evaluating nonlinear terms in the ROMs, a hyper-reduction method via discrete empirical interpolation is applied. Accordingly, we derive error estimates for the ROM approximations of the state and output. Lastly, we demonstrate the structure preservation, as well as the accuracy of the proposed port-Hamiltonian operator inference framework, through numerical experiments on a linear mass–spring-damper problem and a nonlinear Toda lattice problem.

97 MATHEMATICS AND COMPUTING↗

Variational multiscale reinforcement learning for discovering reduced order closure models of nonlinear spatiotemporal transport systems

Abstract A central challenge in the computational modeling and simulation of a multitude of science applications is to achieve robust and accurate closures for their coarse-grained representations due to underlying highly nonlinear multiscale interactions. These closure models are common in many nonlinear spatiotemporal systems to account for losses due to reduced order representations, including many transport phenomena in fluids. Previous data-driven closure modeling efforts have mostly focused on supervised learning approaches using high fidelity simulation data. On the other hand, reinforcement learning (RL) is a powerful yet relatively uncharted method in spatiotemporally extended systems. In this study, we put forth a modular dynamic closure modeling and discovery framework to stabilize the Galerkin projection based reduced order models that may arise in many nonlinear spatiotemporal dynamical systems with quadratic nonlinearity. However, a key element in creating a robust RL agent is to introduce a feasible reward function, which can be constituted of any difference metrics between the RL model and high fidelity simulation data. First, we introduce a multi-modal RL to discover mode-dependant closure policies that utilize the high fidelity data in rewarding our RL agent. We then formulate a variational multiscale RL (VMRL) approach to discover closure models without requiring access to the high fidelity data in designing the reward function. Specifically, our chief innovation is to leverage variational multiscale formalism to quantify the difference between modal interactions in Galerkin systems. Our results in simulating the viscous Burgers equation indicate that the proposed VMRL method leads to robust and accurate closure parameterizations, and it may potentially be used to discover scale-aware closure models for complex dynamical systems.

97 MATHEMATICS AND COMPUTING↗

Parametric conditions for stability of reduced-order linear time-varying control systems

Using a single framework, parametric conditions are derived which encompass those for both local and global BIBO stability of a linear multivariable discrete-time reduced-order time-varying control system. These conditions indicate that the system will be BIBO stable if the norm of the system-parameter error matrix is bounded by an l exp 1 function superimposed on an l exp infinity function.

Ma, C. C. H.↗

The reduced order model problem in distributed parameter systems adaptive identification and control

The basic assumption that a large space structure can be decoupled preceding the application of reduced order active control was considered and alternative solutions to the control of such structures (in contrast to the strict modal control) were investigated. The transfer function matrix from the actuators to the sensors was deemed to be a reasonable candidate. More refined models from multivariable systems theory were studied and recent results in the multivariable control field were compared with respect to theoretical deficiencies and likely problems in application to large space structures.

Johnson, C. R., Jr.↗

A method for high order linear system reduction and nonlinear system simplification

Least-squares-type algorithms for reducing the order of linear systems in the frequency domain and simplifying nonlinear systems in time domain are developed and demonstrated. The possible model structures are represented as nodes in a tree, and costs along the branches are assigned using the repeated-Gram-Schmidt orthogonalization procedure of Desrochers and Saridis (1980), permitting identification of the optimal n-term model by searching the tree to depth n, with no need for parameter identification. The efficiency and flexibility of the algorithms is shown in applications to the eighth-order linear system studied by Hsia (1972), a three-state eight-nonlinear-term aircraft-dynamics problem, and the related linear-controller problem (Garrard and Jordan, 1977).

Desrochers, A. A.↗

A multifidelity deep operator network approach to closure for multiscale systems

Projection-based reduced order models (PROMs) have shown promise in representing the behavior of multiscale systems using a small set of generalized (or latent) variables. Despite their success, PROMs can be susceptible to inaccuracies, even instabilities, due to the improper accounting of the interaction between the resolved and unresolved scales of the multiscale system (known as the closure problem). In the current work, we interpret closure as a multifidelity problem and use a multifidelity deep operator network (DeepONet) framework to address it. In addition, to enhance the stability and/or accuracy of the multifidelity-based closure, we employ the recently developed ``in-the-loop" training approach from the literature on coupling physics and machine learning models. The resulting approach is tested on shock advection for the one-dimensional viscous Burgers equation and vortex merging for the two-dimensional Navier-Stokes equations. Furthermore, the numerical experiments show significant improvement of the predictive ability of the closure-corrected PROM over the un-corrected one both in the interpolative and the extrapolative regimes.

42 ENGINEERING↗

Using integrals of the state transition matrix for efficient transient-response computations

An iterative technique for the computation of lower-order state systems with poles at the origin (such as aircraft lateral dynamics) or with time-polynomial inputs is developed and demonstrated. The integrated input or output signals are computed using the integrals of the state transition matrix, thus reducing the system order and computation time. In an application to the lateral response of a DC-8 aircraft (Grunwald, 1981), this method required only 7824 multiplication or addition operations (at fourth order) to achieve the accuracy obtained in 15,012 operations by a conventional sixth-order computation.

Grunwald, A. J.↗

State reduction for semi-Markov reliability models

Semi-Markov processes have proved to be an effective and convenient tool to construct models of systems that achieve reliability by redundancy and reconfiguration. These models are able to depict complex system architectures and to capture the dynamics of fault arrival and system recovery. A disadvantage of this approach is that the models can be extremely large, which poses both a model construction and a computational problem. Techniques are needed to reduce the model size. Because these systems are used in critical applications where failure can be expensive, there must be an analytically derived bound for the error produced by the model reduction technique. Automatic model generation programs have been written to help the reliability analyst produce models of complex systems. Because of the importance of these programs, the model reduction technique needs to be precise and easily implemented. This paper presents a model reduction technique called trimming that can be applied to a popular class of systems. An error bound for the trimming procedure is derived that uses readily available system parameters. The trimming procedure is precisely described and appears easy to implement in a model generation program.

Reduced order systems↗

Elevation control system model for the DSS 13 antenna

In order to meet the requirements for precision pointing of 34-m antennas, adequate control design and simulation software have to be developed along with a detailed description of the supporting analytical tools. This article describes a control system model for the elevation drive of the DSS 13 antenna. The model allows one to simulate elevation dynamics, cross-coupled dynamics in azimuth and elevation, and RF pointing error. A modal state-space model of the antenna structure was obtained from its finite-element model with a free rotating tipping structure. Model reduction techniques were applied separately for the antenna model and rate-loop model, thereby reducing the system order to one-third of the original one while preserving its dynamic properties. Extensive simulation results illustrate properties of the model.

Gawronski, W.↗

Mechanical Testing of PMCs under Simulated Rapid Heat-Up Propulsion Environments: In-Plane Compressive Behavior - II

Carbon fiber thermoset polymer matrix composites (PMC) with high temperature polyimide based in-situ polymerized monomer reactant (PMR) resin has been used for some time in applications which can see temperatures up to 550 F. Currently, graphite fiber PMR based composites are used in several aircraft engine components including the outer bypass duct for the GE F-404, exit flaps for the P&W F-100-229, and the core cowl for the GE/Snecma CF6-80A3. Newer formulations, including PMR-II-50 are being investigated as potential weight reduction replacements of various metallic components in next generation high performance propulsion rocket engines that can see temperatures which exceed 550 F. Extensive FEM thermal modeling indicates that these components are exposed to rapid heat-up rates (up to -200 F/sec) and to a maximum temperature of around 600 F. Even though the predicted maximum part temperatures were within the capability of PW-II-50, the rapid heat-up causes significant through-thickness thermal gradients in the composite part and even more unstable states when combined with moisture. Designing composite parts for such extreme service environments will require accurate measurement of intrinsic and transient mechanical properties and the hygrothermal performance of these materials under more realistic use conditions. The mechanical properties of polymers degrade when exposed to elevated temperatures even in the absence of gaseous oxygen. Accurate mechanical characterization of the material is necessary in order to reduce system weight while providing sufficient factors of safety. Historically, the testing of PMCs at elevated temperatures has been plagued by the antagonism between two factors. First, moisture has been shown to profoundly affect the mechanical response of these materials at temperatures above their glass transition temperature while concurrently lowering the material's Tg. Moisture phenomena is due to one or a combination of three effects, i.e., plastization of polymeric material by water, the internal pressure generated by the volatilization of water at elevated temperatures, and hydrolytic chemical decomposition. However, moisture is lost from the material at increasing rates as temperature increases. Second, because PMCs are good thermal insulators, when they are externally heated at even mild rates large thermal gradients can develop within the material. At temperatures where a material property changes rapidly with temperature the presence of a large thermal gradient is unacceptable for intrinsic property characterization purposes. Therefore, long hold times are required to establish isothermal conditions. However, in the service environments high-heating-rates, high temperatures, high-loading rates are simultaneous present along with residual moisture. In order to capture the effects of moisture on the material, holding at- temperature until isothermal conditions are reached is unacceptable particularly in materials with small physical dimensions. Thus, the effects due to moisture on the composite's mechanical characteristics, ie., their so-called analog response, may be instructive. One approach employed in this program was rapid heat-up (approx. 200 F/sec.) and loading of both dry and wet in-plane compressive specimens to examine the effects of moisture on this resin dominated mechanical property of the material.

Stokes, Eric H.↗

Influences of joystick spring resistance on the execution of simple and complex positioning movements

To provide good proprioceptive feedback in a manual control device for a designation task, spring resistance of a joystick was optimized by adjustment of centering force and deflection nonlinearly with each other by using the psychophysical method of cross modality matching. Designation with zero and first order systems showed that the coarse adjustment was insensitive to stick and certain task parameters, although it was influenced by some biomechanical parameters and the anticipated demands of the final control positioning. Only the more difficult fine adjustment is sensitive to parameter alterations and therefore suitable for optimization attempts. The strong centering of the stick by a nonlinear degressive spring resistance facilitates fine adjustment. Through this, total adjustment time with the first order system is reduced by more than thirty percent, compared to a linear resistance. Tracking experiments affirm the usefulness and preference of nonlinear spring resistance.

Rothbauer, G.↗

Model reduction for discrete bilinear systems

A model reduction method for discrete bilinear systems is developed which matches q sets of Volterra and covariance parameters. These parameters are shown to represent both deterministic and stochastic attributes of the discrete bilinear system. A reduced order model which matches these q sets of parameters is defined to be a q-Volterra covariance equivalent realization (q-Volterra COVER). An algorithm is presented which constructs a class of q-Volterra COVERs parameterized by solutions to a Hermitian, quadratic, matrix equation. The algorithm is applied to a bilinear model of a robot manipulator.

King, A. M.↗

Modeling and simulations of the DSS 13 antenna control system

A model of the antenna control system for the azimuth and elevation axes of the Deep Space Station (DSS) 13 antenna is developed. This model is used for simulation of elevation and azimuth dynamics, cross-coupled dynamics, and radio-frequency pointing error due to both input commands and wind disturbances. This model also serves as a tool for the antenna controller design. A modal state-space model of the antenna structure was obtained from its finite-element model with a free-rotating tipping structure and alidade. Model reduction techniques applied separately for the antenna structure, elevation, and azimuth drives, and rate-loop model reduce the system order to one-third of that of the original, while preserving its significant dynamic properties. Extensive simulation results illustrate properties of the model.

Gawronski, W.↗

Discrete-time model reduction in limited frequency ranges

A mathematical formulation for model reduction of discrete time systems such that the reduced order model represents the system in a particular frequency range is discussed. The algorithm transforms the full order system into balanced coordinates using frequency weighted discrete controllability and observability grammians. In this form a criterion is derived to guide truncation of states based on their contribution to the frequency range of interest. Minimization of the criterion is accomplished without need for numerical optimization. Balancing requires the computation of discrete frequency weighted grammians. Close form solutions for the computation of frequency weighted grammians are developed. Numerical examples are discussed to demonstrate the algorithm.

Horta, Lucas G.↗