Search NASA⌕ Search

SEARCH · Search NASA

Results for “robust parameter design”

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 253 records · Page 14

Validating an Air Traffic Management Concept of Operation Using Statistical Modeling

Validating a concept of operation for a complex, safety-critical system (like the National Airspace System) is challenging because of the high dimensionality of the controllable parameters and the infinite number of states of the system. In this paper, we use statistical modeling techniques to explore the behavior of a conflict detection and resolution algorithm designed for the terminal airspace. These techniques predict the robustness of the system simulation to both nominal and off-nominal behaviors within the overall airspace. They also can be used to evaluate the output of the simulation against recorded airspace data. Additionally, the techniques carry with them a mathematical value of the worth of each prediction-a statistical uncertainty for any robustness estimate. Uncertainty Quantification (UQ) is the process of quantitative characterization and ultimately a reduction of uncertainties in complex systems. UQ is important for understanding the influence of uncertainties on the behavior of a system and therefore is valuable for design, analysis, and verification and validation. In this paper, we apply advanced statistical modeling methodologies and techniques on an advanced air traffic management system, namely the Terminal Tactical Separation Assured Flight Environment (T-TSAFE). We show initial results for a parameter analysis and safety boundary (envelope) detection in the high-dimensional parameter space. For our boundary analysis, we developed a new sequential approach based upon the design of computer experiments, allowing us to incorporate knowledge from domain experts into our modeling and to determine the most likely boundary shapes and its parameters. We carried out the analysis on system parameters and describe an initial approach that will allow us to include time-series inputs, such as the radar track data, into the analysis

Statistical emulation↗

Robust Control Design for Uncertain Nonlinear Dynamic Systems

Robustness to parametric uncertainty is fundamental to successful control system design and as such it has been at the core of many design methods developed over the decades. Despite its prominence, most of the work on robust control design has focused on linear models and uncertainties that are non-probabilistic in nature. Recently, researchers have acknowledged this disparity and have been developing theory to address a broader class of uncertainties. This paper presents an experimental application of robust control design for a hybrid class of probabilistic and non-probabilistic parametric uncertainties. The experimental apparatus is based upon the classic inverted pendulum on a cart. The physical uncertainty is realized by a known additional lumped mass at an unknown location on the pendulum. This unknown location has the effect of substantially altering the nominal frequency and controllability of the nonlinear system, and in the limit has the capability to make the system neutrally stable and uncontrollable. Another uncertainty to be considered is a direct current motor parameter. The control design objective is to design a controller that satisfies stability, tracking error, control power, and transient behavior requirements for the largest range of parametric uncertainties. This paper presents an overview of the theory behind the robust control design methodology and the experimental results.

Kenny, Sean P.↗

MaLTESE: Large-Scale Simulation-Driven Machine Learning for Transient Driving Cycles

Optimal engine operation during a transient driving cycle is the key to achieving greater fuel economy, engine efficiency, and reduced emissions. In order to achieve continuously optimal engine operation, engine calibration methods use a combination of static correlations obtained from dynamometer tests for steady-state operating points and road and/or track performance data. As the parameter space of control variables, design variable constraints, and objective functions increases, the cost and duration for optimal calibration become prohibitively large. In order to reduce the number of dynamometer tests required for calibrating modern engines, a large-scale simulation-driven machine learning approach is presented in this work. A parallel, fast, robust, physics-based reduced-order engine simulator is used to obtain performance and emission characteristics of engines over a wide range of control parameters under various transient driving conditions (drive cycles). We scale the simulation up to 3,906 nodes of the Theta supercomputer at the Argonne Leadership Computing Facility to generate data required to train a machine learning model. The trained model is then used to predict various engine parameters of interest, and the results are compared with those predicted by the engine simulator. Our results show that a deepneural-network-based surrogate model achieves high accuracy: Pearson product-moment correlation values larger than 0.99 and mean absolute percentage error within 1.07% for various engine parameters such as exhaust temperature, exhaust pressure, nitric oxide, and engine torque. Once trained, the deep-neural-network-based surrogate model is fast for inference: it requires about 16 mu s for predicting the engine performance and emissions for a single design configuration compared with about 0.5 s per configuration with the engine simulator. Moreover, we demonstrate that transfer learning and retraining can be leveraged to incrementally retrain the surrogate model to cope with new configurations that fall outside the training data space.

deep learning↗

Robust controllers for the Middeck Active Control Experiment using Popov controller synthesis

Recent work in robust control with real parameter uncertainties has focused on absolute stability and its connections to real mu theory. In particular, the research has investigated the Popov stability criterion and its associated Lur'e-Postnikov Liapunov functions. State space representations of this Popov stability analysis tests are included in an H2 design formulation to provide a powerful technique for robust controller synthesis. This synthesis approach uses a state space optimization procedure to design controllers that minimize an overbound of an H2 cost functional and satisfy stability analysis tests based on the Popov multiplier. The controller and stability multiplier coefficients are optimized simultaneously, which avoids the iteration and curve-fitting procedures required by the D-K algorithm of mu synthesis. While previous work has demonstrated this synthesis approach on benchmark control problems, the purpose of this paper is to use Popov controller synthesis to design robust compensators for the Middeck Active Control Experiment (MACE).

How, Jonathan P.↗

Active control synthesis for flexible space structures excited by persistent disturbances

Both classical and state-space synthesis methods for active control of flexible space structures in the presence of persistent disturbances are presented. The methods exploit the so-called internal model principle for asymptotic disturbance rejection. A generic example of flexible space structures is used to illustrate the simplicity of the proposed design methodologies. The concept of a disturbance rejection filter dipole is introduced from a classical control viewpoint. It is shown that the proposed design methods will invariably make use of non-minimum-phase compensation for a class of noncolocated control problems. The need for tradeoffs between performance and parameter robustness is discussed.

Wie, Bong↗

Semigroup approximation and robust stabilization of distributed parameter systems

Theoretical results that enable rigorous statements of convergence and exponential stability of Galerkin approximations of LQR controls for infinite dimensional, or distributed parameter, systems have proliferated over the past ten years. In addition, extensive progress has been made over the same time period in the derivation of robust control design strategies for finite dimensional systems. However, the study of the convergence of robust finite dimensional controllers to robust controllers for infinite dimensional systems remains an active area of research. We consider a class of soft-constrained differential games evolving in a Hilbert space. Under certain conditions, a saddle point control can be given in feedback form in terms of a solution to a Riccati equation. By considering a related LQR problem, we can show a convergence result for finite dimensional approximations of this differential game. This yields a computational algorithm for the feedback gain that can be derived from similar strategies employed in infinite dimensional LQR control design problems. The approach described in this paper also inherits the additional properties of stability robustness common to game theoretic methods in finite dimensional analysis. These theoretical convergence and stability results are verified in several numerical experiments.

Kurdila, A. J.↗

Robust and real-time control of magnetic bearings for space engines

Currently, NASA Lewis Research Center is developing magnetic bearings for Space Shuttle Main Engine (SSME) turbopumps. The control algorithms which have been used are based on either the proportional-intergral-derivative control (PID) approach or the linear quadratic (LQ) state space approach. These approaches lead to an acceptable performance only when the system model is accurately known, which is seldom true in practice. For example, the rotor eccentricity, which is a major source of vibration at high speeds, cannot be predicted accurately. Furthermore, the dynamics of a rotor shaft, which must be treated as a flexible system to model the elastic rotor shaft, is infinite dimensional in theory and the controller can only be developed on the basis of a finite number of modes. Therefore, the development of the control system is further complicated by the possibility of closed loop system instability because of residual or uncontrolled modes, the so called spillover problem. Consequently, novel control algorithms for magnetic bearings are being developed to be robust to inevitable parametric uncertainties, external disturbances, spillover phenomenon and noise. Also, as pointed out earlier, magnetic bearings must exhibit good performance at a speed over 30,000 rpm. This implies that the sampling period available for the design of a digital control system has to be of the order of 0.5 milli-seconds. Therefore, feedback coefficients and other required controller parameters have to be computed off-line so that the on-line computational burden is extremely small. The development of the robust and real-time control algorithms is based on the sliding mode control theory. In this method, a dynamic system is made to move along a manifold of sliding hyperplanes to the origin of the state space. The number of sliding hyperplanes equals that of actuators. The sliding mode controller has two parts; linear state feedback and nonlinear terms. The nonlinear terms guarantee that the systems would reach the intersection of all sliding hyperplanes and remain on it when bounds on the errors in the system parameters and external disturbances are known. The linear part of the control drives the system to the origin of state space. Another important feature is that the controller parameter can be computed off-line. Consequently, on-line computational burden is small.

Sinha, Alok↗

Robust Nonlinear Feedback Control of Aircraft Propulsion Systems

This is the final report on the research performed under NASA Glen grant NASA/NAG-3-1975 concerning feedback control of the Pratt & Whitney (PW) STF 952, a twin spool, mixed flow, after burning turbofan engine. The research focussed on the design of linear and gain-scheduled, multivariable inner-loop controllers for the PW turbofan engine using H-infinity and linear, parameter-varying (LPV) control techniques. The nonlinear turbofan engine simulation was provided by PW within the NASA Rocket Engine Transient Simulator (ROCETS) simulation software environment. ROCETS was used to generate linearized models of the turbofan engine for control design and analysis as well as the simulation environment to evaluate the performance and robustness of the controllers. Comparison between the H-infinity, and LPV controllers are made with the baseline multivariable controller and developed by Pratt & Whitney engineers included in the ROCETS simulation. Simulation results indicate that H-infinity and LPV techniques effectively achieve desired response characteristics with minimal cross coupling between commanded values and are very robust to unmodeled dynamics and sensor noise.

Garrard, William L.↗

Chaos control of the brushless direct current motor using adaptive dynamic surface control based on neural network with the minimum weights

This paper investigates chaos control for the brushless DC motor (BLDCM) system by adaptive dynamic surface approach based on neural network with the minimum weights. The BLDCM system contains parameter perturbation, chaotic behavior, and uncertainty. With the help of radial basis function (RBF) neural network to approximate the unknown nonlinear functions, the adaptive law is established to overcome uncertainty of the control gain. By introducing the RBF neural network and adaptive technology into the dynamic surface control design, a robust chaos control scheme is developed. It is proved that the proposed control approach can guarantee that all signals in the closed-loop system are globally uniformly bounded, and the tracking error converges to a small neighborhood of the origin. Simulation results are provided to show that the proposed approach works well in suppressing chaos and parameter perturbation.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Compressive properties of silicone Bouligand structures

Abstract This article presents an examination of silicone, Bouligand lattices in compression. Appearing frequently in biological organisms and manufacturing design, Bouligand structures comprise layers of parallel strands or fibers oriented in a helical fashion. They can exhibit exceptional fracture resistance when composed of rigid or composite materials. The behavior of elastomeric Bouligand structures, however, is less well understood. Additively manufactured (AM) elastomeric lattices have applications in stress mitigation, medical devices, and soft robotics. This article demonstrates that Bouligand structures are a useful addition to the design space of AM elastomers. By adjusting the layer-rotation parameters, lattice stress can increase by more than 300% without altering the porosity. Additionally, we introduce path length metrics that help explain the observed relationship between layer rotation and compression response. Impact statement Additive manufacturing (AM) continues to push the boundary of manufacturable structures and enhance the ability to robustly design for specific properties and behaviors. The more we understand the design space of a novel AM microstructure, the greater its application range. In this article, we describe the mechanical behavior of helicoidal, elastomeric lattices and introduce path length metrics to help explain their stress response. We show that these structures can exhibit a large range of mechanical behaviors in compression, making them well suited for applications such as stress mitigation and impact absorption. Additionally, the path length metrics could become useful design tools and may be applicable to a larger set of cellular structures. These findings expand our ability to rapidly design materials with highly specific and customizable properties to meet the needs of modern engineering challenges. Graphical abstract

Van Meerbeek, I. M. (ORCID:000000021258132X)↗

Implementation of a Model Predictive Control Strategy to Regulate Temperature Inside Plug-Flow Solar Reactor With Countercurrent Flow

Abstract Solar-driven thermochemical energy storage systems are proven to be promising energy carriers (solar fuels) to utilize solar energy by using reactive solid-state pellets. However, the production of solar fuel requires a quasi-steady-state process temperature, which represents the main challenge due to the transient nature of solar power. In this work, an adaptive model predictive controller (MPC) is presented to regulate the temperature inside a tubular solar reactor to produce solid-state solar fuel for long-term thermal storage systems. The solar reactor system consists of a vertical tube heated circumferentially over a segment of its length by concentrated solar power, and the reactive pellets (MgMn2O4) are fed from the top end and flow downwards through the heated tube. A countercurrent flowing gas supplied from the lower end interacts with flowing pellets to reduce it thermochemically at a temperature range of 1000—1500 °C. A low-order physical model was developed to simulate the dynamics of the solar reactor including the reaction kinetics, and the proposed model was validated numerically by using a 7-kW electric furnace. The numerical model then was utilized to design the MPC controller, where the control system consists of an MPC code linked to an adaptive system identification code that updates system parameters online to ensure system robustness against external disturbances (sudden change in the flow inside the reactor), model mismatches, and uncertainty. The MPC controller parameters are tuned to enhance the system performance with minimum steady-state error and overshoot. The controller is tested to track different temperature ranges between 500 °C and 1400 °C with different particles/gas mass flowrates and ramping temperature profiles. Results show that the MPC controller successfully regulated the reactor temperature within ± 1 °C of its setpoint and maintained robust performance with minimum input effort when subjected to sudden changes in the amount of flowing media and the presence of chemical reaction.

Engineering↗

Coupling of CTF and TRANSFORM using the Functional Mockup Interface

An in-memory coupling between the sub-channel thermal hydraulics code COBRA-TF (CTF), which is included in the Virtual Environment for Reactor Applications (VERA), and the systems code Transient Simulation Framework of Reconfigurable Models (TRANSFORM) was developed in this work. Data exchange is accomplished by using the Functional Mock-Up Interface (FMI), an open standard for coupling dynamic systems models together. The FMI-based coupling necessitated the development of a novel FORTRAN wrapper for communicating with Functional Mock-Up Units. This wrapper will facilitate future FMI-based code couplings with other FORTRAN-based programs. CTF-TRANSFORM is exercised on a simplified Molten Salt Reactor Experiment (MSRE) model in steady-state and transient configurations. The coupled CTF-TRANSFORM model is shown to predict core temperature deltas similar to those available in historical MSRE operational data and is shown to be robust to fast power transients.The coupling paves the way to performing sensitivity and design studies impossible with VERA/CTF alone, in which secondary and tertiary loop operating conditions and design parameters may be perturbed and their impact on the core operating conditions studied.

97 MATHEMATICS AND COMPUTING↗

Development of An Intelligent Flight Propulsion Control System

The initial design and demonstration of an Intelligent Flight Propulsion and Control System (IFPCS) is documented. The design is based on the implementation of a nonlinear adaptive flight control architecture. This initial design of the IFPCS enhances flight safety by using propulsion sources to provide redundancy in flight control. The IFPCS enhances the conventional gain scheduled approach in significant ways: (1) The IFPCS provides a back up flight control system that results in consistent responses over a wide range of unanticipated failures. (2) The IFPCS is applicable to a variety of aircraft models without redesign and,(3) significantly reduces the laborious research and design necessary in a gain scheduled approach. The control augmentation is detailed within an approximate Input-Output Linearization setting. The availability of propulsion only provides two control inputs, symmetric and differential thrust. Earlier Propulsion Control Augmentation (PCA) work performed by NASA provided for a trajectory controller with pilot command input of glidepath and heading. This work is aimed at demonstrating the flexibility of the IFPCS in providing consistency in flying qualities under a variety of failure scenarios. This report documents the initial design phase where propulsion only is used. Results confirm that the engine dynamics and associated hard nonlineaaities result in poor handling qualities at best. However, as demonstrated in simulation, the IFPCS is capable of results similar to the gain scheduled designs of the NASA PCA work. The IFPCS design uses crude estimates of aircraft behaviour. The adaptive control architecture demonstrates robust stability and provides robust performance. In this work, robust stability means that all states, errors, and adaptive parameters remain bounded under a wide class of uncertainties and input and output disturbances. Robust performance is measured in the quality of the tracking. The results demonstrate the flexibility of the IFPCS architecture and the ability to provide robust performance under a broad range of uncertainty. Robust stability is proved using Lyapunov like analysis. Future development of the IFPCS will include integration of conventional control surfaces with the use of propulsion augmentation, and utilization of available lift and drag devices, to demonstrate adaptive control capability under a greater variety of failure scenarios. Further work will specifically address the effects of actuator saturation.

Calise, A. J.↗

Robustness of inertial fusion energy relevant implosions to low-mode asymmetries

Inertial confinement fusion implosions for power production need to be robust. Degradation sources need to be understood and considered for an economic power plant. In this work, the impact of low-mode asymmetries on yield is studied using 2D radiation hydrodynamics simulations for inertial fusion energy relevant implosions. First, a high-gain indirectly driven capsule design is developed. This design is subsequently degraded with low-mode drive asymmetries. The results show that yield is constant with increasing amounts of asymmetry until a cliff is reached where the yield quickly drops below ignition. Robustness arises because the degraded hotspot still has enough time to get to ignition before capsule decompression. This robustness occurs even though the asymmetry couples less energy to the hotspot and causes decompression to occur sooner. The robustness is linked to the generalized Lawson parameter, χ⁠. The value of χ steadily decreases with increasing asymmetry. Once χ is close to one, the hotspot is no longer able to ignite before the capsule decompresses. A simple piston model is used to predict where the robustness cliff is for any design. The trade-off between yield and robustness is also demonstrated.

Physics - Plasma physics↗

Control, Fault Management, and Grid Support Functionality of an MV AC-DC Solid State Transformer based EV Extreme Fast Charging Station

Electric vehicles (EVs) have become increasingly popular in recent times while revolutionizing the consumer and commercial transportation market. The development of charging infrastructure has become one of the priorities for increasing the adoption of EVs. Extreme fast charging (XFC) technology can reduce the so-called ’range anxiety’ of consumers as they significantly reduce the charging time. With the advent of wide band-gap (WBG) power devices and improvement in power electronic converters, medium voltage (MV) solid state transformer (SST) based XFC system has the potential to replace the traditional XFC stations because of the lower footprint, ease of installation, enhanced control feature, and better system efficiency. The control system design is one of the critical aspects of the SST development process. Careful consideration and detailed analysis are required to find out suitable control method for the SST based on its topology among different centralized and decentralized control architectures. Also, the control parameters selection and potential improvement to the transient response of the controller ought to be investigated. Another major concern of the SST is different types of internal fault which reduces the overall reliability of the XFC system. As a result, designing a robust protection system is essential. Among different fault modes, open circuit switch faults have received significant attention as an active research area because of their likelihood and severe effects on converters. Therefore, the power stages used in the XFC system require functional and accurate open circuit switch fault management methods. An equally significant aspect of this SST based XFC is its compatibility in a microgrid where there is no synchronous generator present. When the grid is not available, the XFC SSTs can provide grid forming capability and continue supplying the critical loads in islanded mode. The transition between grid connected and islanded mode, especially the grid resynchronization process has to be carefully performed for the safety of the microgrid components. The challenges posed by the aforementioned issues have inspired the work done in this dissertation. Here, a 13.2 kV, 1 MVA, AC/DC SST for the XFC system is examined and a comparative analysis is conducted to select the control architecture based on feasibility of implementation and performance. A detailed control parameter design process is demonstrated considering the sensor dynamics and delay. The selected decentralized control method is augmented by introducing a novel sensor-less load current feedforward method to provide better voltage regulation at the DC bus during a change of load. Next, in the fault management section, a hierarchical failure mode effect analysis (FMEA) is proposed to enable a systematic design of the internal fault protection of the XFC SST as there are limited examples in the literature regarding the analysis of the safety and design of the protection of a power electronic converter system. Novel open circuit switch fault management methods for the converters in the system are presented. Finally, XFC SST based MV microgrid operations in grid connected mode and islanded mode are explored. A secondary control method for grid resynchronization is presented and a design process of control parameters is shown to ensure the stability of the secondary voltage and frequency regulation.

30 DIRECT ENERGY CONVERSION↗

Evaluation of Tungsten—Steel Solid-State Bonding: Options and the Role of CALPHAD to Screen Diffusion Bonding Interlayers

Critical aspects of innovative design in engineering disciplines like infrastructure, transportation, and medical applications require the joining of dissimilar materials. This study investigates the literature on solid-state bonding techniques, with a particular focus on diffusion bonding, as an effective method for establishing engineering bonds. Welding and brazing, while widely used, may pose challenges when joining materials with large differences in melting temperature and can lead to mechanical property degradation. In contrast, diffusion bonding offers a lower temperature process that relies on solid-state interactions to develop bond strength. The joining of tungsten and steel, especially for fusion reactors, presents a unique challenge due to the significant disparity in melting temperatures and the propensity to form brittle intermetallics. Here, diffusion characteristics of tungsten–steel interfaces are examined and the influence of bonding parameters on mechanical properties are investigated. Additionally, CALPHAD modeling is employed to explore joining parameters, thermal stability, and diffusion kinetics. The insights from this research can be extended to join numerous dissimilar materials for specific applications such as aerospace, automobile industry, power plants, etc., enabling advanced and robust design with high efficiency.

36 MATERIALS SCIENCE↗

Secure Federated Learning Across Heterogeneous Cloud and High-Performance Computing Resources: A Case Study on Federated Fine-Tuning of LLaMA 2

Federated learning enables multiple data owners to collaboratively train robust machine learning models without transferring large or sensitive local datasets by only sharing the parameters of the locally trained models. Here, in this article, we elaborate on the design of our Advanced Privacy-Preserving Federated Learning (APPFL) framework, which streamlines end-to-end secure and reliable federated learning experiments across cloud computing facilities and high-performance computing resources by leveraging Globus Compute, a distributed function as a service platform, and Amazon Web Services. We further demonstrate the use case of APPFL in fine-tuning an LLaMA 2 7B model using several cloud resources and supercomputers.

97 MATHEMATICS AND COMPUTING↗

Novel Compact Ceilometer (Final Scientific-Technical Report)

Fibertek is developing a novel ceilometer system that provides a new measurement capability not currently available to weather stations such as those in the Department of Energy's AmeriFlux network. Commercially available instruments do not have the sensitivity to see atmospheric aerosols. If the atmospheric aerosols can be measured, then the altitude of the boundary layer can be determined from a discontinuity in the aerosol concentration. The altitude of the boundary layer is a critical parameter in climate and weather models, and adding this piece of information will greatly improve the accuracy of these models. Fibertek's design provides this capability in an inexpensive robust fieldable instrument by applying a new sensor technology: the Silicon Photomultiplier (SiPM). SiPMs are called "photomultipliers" in that they provide very similar performance as a traditional photomultiplier including single photon sensitivity and linear response. However, they are a completely different technology which offers several advantages. They are small, inexpensive, solid-state, and extremely robust. They do not require cooling. They operate at much lower voltages. They cannot be damaged by inadvertent reflections, and when saturated they recover in less than a microsecond, so they don't need to be shielded from the outgoing laser pulse. In short, they make single photon sensitivity accessible for an inexpensive, simple, rugged, fieldable system. In this Phase 1 effort Fibertek developed a breadboard version of the ceilometer to demonstrate the feasibility of our approach. Our partner, Penn State, conducted atmospheric measurements to validate the lidar’s performance and determine algorithms for automatically identifying the height of the boundary layer. The lidar shows sufficient sensitivity to fully characterize the boundary layer in clear air daytime and nighttime conditions.

47 OTHER INSTRUMENTATION↗