Search NASA⌕ Search

SEARCH · Search NASA

Results for “model predictive controller”

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 901 records · Page 50

Neural Network Analysis of Nuclear Magnetic Resonance and Infrared Spectra

Nuclear magnetic resonance (NMR) spectroscopy and infrared (IR) spectroscopy are powerful chemical characterization techniques with broad general usage. However, the manual evaluation of the resulting spectra is time-consuming and requires significant expertise, preventing insights from being used in real-time applications. With recent advances in computation and artificial intelligence (AI), new tools are available for automating spectral interpretation. In this work, machine learning (ML) algorithms using 1-dimensional convolutional neural networks (CNNs) were applied to identify common functional groups from spectral information. Raw spectra were collected virtually from the Human Metabolome Database (HMDB) and National Institute of Standards and Technology (NIST) Chemistry WebBook and processed into a suitable standard. Algorithm design was tailored to best fit the nature of the problem, with built-in flexibility to accommodate relevant parameters beyond the raw spectral input, specifically solvent identity and magnetic frequency for NMR. The predictive capability of the algorithm in identifying functional groups is displayed in several examples. This methodology has been compiled into a code repository and could easily be modified to adapt alternative data sources, including other spectrum types. To mitigate overfitting, a common problem in mathematical modeling where overfamiliarity with training data produces trends that are not representative of the general data, a novel metric was developed, referred to as Accufit. Accufit includes a parameter that penalizes substantial differences in the training accuracy and the accuracy of an independent validation set. Examples are presented showing the effectiveness of Accufit in maintaining the model’s predictive capability while controlling the overfitting when used as a custom metric for hyperparameter tuning.

Sturgill, James↗

Machine learning-enhanced model-based scenario optimization for DIII-D

Abstract Scenario development in tokamaks is an open area of investigation that can be approached in a variety of different ways. Experimental trial and error has been the traditional method, but this required a massive amount of experimental time and resources. As high fidelity predictive models have become available, offline development and testing of proposed scenarios has become an option to reduce the required experimental resources. The use of predictive models also offers the possibility of using a numerical optimization process to find the controllable inputs that most closely achieve the desired plasma state. However, this type of optimization can require as many as hundreds or thousands of predictive simulation cases to converge to a solution; many of the commonly used high fidelity models have high computational burdens, so it is only reasonable to run a handful of predictive simulations. In order to make use of numerical optimization approaches, a compromise needs to be found between model fidelity and computational burden. This compromise can be achieved using neural networks surrogates of high fidelity models that retain nearly the same level of accuracy as the models they are trained to replicate while reducing the computation time by orders of magnitude. In this work, a model-based numerical optimization tool for scenario development is described. The predictive model used by the optimizer includes neural network surrogate models integrated into the fast Control-Oriented Transport simulation framework. This optimization scheme is able to converge to the optimal values of the controllable inputs that produce the target plasma scenario by running thousands of predictive simulations in under an hour without sacrificing too much prediction accuracy.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Controlling Microbial Byproducts using Model-Based Substrate Monitoring and Control Strategies

We have developed a computer-controlled bioreactor system to study various aspects of microbially-mediated nitrogen cycling. The system has been used to investigate methods for controlling microbial denitrification (the dissimilatory reduction of nitrate to N2O and N2) in hydroponic plant growth chambers. Such chambers are key elements of advanced life support systems being designed for use on long duration space missions, but nitrogen use efficiency in them is reduced by denitrification. Control software architecture was designed which permits the heterogeneous control of system hardware using traditional feedback control, and quantitative and qualitative models of various system features. Model-based feed forward control entails prediction of future systems in states and automated regulation of system parameters to achieve desired and avoid undesirable system states. A bacterial growth rate model based on the classic Monod model of saturation kinetics was used to evaluate the response of several individual denitrifying species to varying environmental conditions. The system and models are now being applied to mixed microbial communities harvested from the root zone of a hydroponic growth chamber. The use of a modified Monod organism interaction model was evaluated as a means of achieving more accurate description of the dynamic behavior of the communities. A minimum variance parameter estimation routine was also' used to calibrate the constant parameters in the model by iterative evaluation of substrate (nitrate) uptake and growth kinetics. This representation of processes and interactions aids in the formulation of control laws. The feed forward control strategy being developed will increase system autonomy, reduce crew intervention and limit the accumulation of undesirable waste products (NOx).

Smernoff, David T.↗

Controlling bimetallic nanostructures by the microemulsion method with subnanometer resolution using a prediction model

In this study, we present a theoretical model to predict the atomic structure of Au/Pt nanoparticles synthesized in microemulsions. Excellent concordance with the experimental results shows that the structure of the nanoparticles can be controlled at sub-nanometer resolution simply by changing the reactants concentration. The results of this study not only offer a better understanding of the complex mechanisms governing reactions in microemulsions, but open up a simple new way to synthesize bimetallic nanoparticles with ad-hoc controlled nanostructures.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

A control-theory model for human decision-making

A model for human decision making is an adaptation of an optimal control model for pilot/vehicle systems. The models for decision and control both contain concepts of time delay, observation noise, optimal prediction, and optimal estimation. The decision making model was intended for situations in which the human bases his decision on his estimate of the state of a linear plant. Experiments are described for the following task situations: (a) single decision tasks, (b) two-decision tasks, and (c) simultaneous manual control and decision making. Using fixed values for model parameters, single-task and two-task decision performance can be predicted to within an accuracy of 10 percent. Agreement is less good for the simultaneous decision and control situation.

Levison, W. H.↗

Predictive modeling and reducing cyclic variability in autoignition engines

Methods and systems are provided for controlling a vehicle engine to reduce cycle-to-cycle combustion variation. A predictive model is applied to predict cycle-to-cycle combustion behavior of an engine based on observed engine performance variables. Conditions are identified, based on the predicted cycle-to-cycle combustion behavior, that indicate high cycle-to-cycle combustion variation. Corrective measures are then applied to prevent the predicted high cycle-to-cycle combustion variation.

33 ADVANCED PROPULSION SYSTEMS↗

Predictive Thermal Control (PTC) Technology to Enable Thermally Stable Telescopes

The Predictive Thermal Control (PTC) technology development project is a multiyear effort initiated in Fiscal Year (FY) 2017, to mature the Technology Readiness Level (TRL) of critical technologies required to enable ultra-thermally-stable telescopes for exoplanet science. During 2017/18 PTC has successfully progressed its three defined objectives: 1. Validate thermal optical performance models. 2. Derive thermal system stability specifications. 3. Demonstrate Predictive Thermal Control. by accomplishing or advancing three of its five quantifiable milestones: Milestone #1 (Complete): Created a high-fidelity STOP (Structural-Thermal-Optical-Performance) model of the1.5 meter ULE® AMTD (Advanced Mirror Technology Development)-2 mirror, including 3D CTE (Coefficient of Thermal Expansion) distribution and reflective coating, that predicts its optical performance response to steady-state and dynamic thermal gradients. Structural model was created using 3D X-Ray Computed Tomography; Milestone #2 (Complete): Derived specifications for thermal control system as a function of wavefront stability for a Vector Vortex Coronagraph; Milestone #3 (In-Process): Designed and started fabricating a predictive Thermal Control System for the AMTD-2 1.5 meter ULE® mirror that senses temperature changes and actively controls the mirror's thermal environment; Milestone #4 (In-Process): Performed preliminary STOP model validation tests of the 1.5-m ULE® mirror in a relevant thermal vacuum environment at the MSFC X-ray and Cryogenic Facility (XRCF) test facility; Milestone #5: Use validated model to perform trade studies to determine how thermo-optical performance can be optimized as a function of mirror design, material selection, mass, etc.

Stahl, H. Philip↗

Toward Optimal Transport Networks

Strictly evolutionary approaches to improving the air transport system a highly complex network of interacting systems no longer suffice in the face of demand that is projected to double or triple in the near future. Thus evolutionary approaches should be augmented with active design methods. The ability to actively design, optimize and control a system presupposes the existence of predictive modeling and reasonably well-defined functional dependences between the controllable variables of the system and objective and constraint functions for optimization. Following recent advances in the studies of the effects of network topology structure on dynamics, we investigate the performance of dynamic processes on transport networks as a function of the first nontrivial eigenvalue of the network's Laplacian, which, in turn, is a function of the network s connectivity and modularity. The last two characteristics can be controlled and tuned via optimization. We consider design optimization problem formulations. We have developed a flexible simulation of network topology coupled with flows on the network for use as a platform for computational experiments.

Alexandrov, Natalia↗

System identification of the Large-Angle Magnetic Suspension Test Facility (LAMSTF)

The Large-Angle Magnetic Suspension Test Facility (LAMSTF), a laboratory-scale research project to demonstrate the magnetic suspension of objects over wide ranges of attitudes, has been developed. This system represents a scaled model of a planned Large-Gap Magnetic Suspension System (LGMSS). The LAMSTF system consists of a planar array of five copper electromagnets which actively suspend a small cylindrical permanent magnet. The cylinder is a rigid body and can be controlled to move in five independent degrees of freedom. Five position variables are sensed indirectly by using infra-red light-emitting diodes and light-receiving phototransistors. The motion of the suspended cylinder is in general nonlinear and hence only the linear, time-invariant perturbed motion about an equilibrium state is considered. One of the main challenges in this project is the control of the suspended element over a wide range of orientations. An accurate dynamic model plans an essential role in controller design. The analytical model of the LAMSTF system includes highly unstable real poles (about 10 Hz) and low-frequency flexible modes (about 0.16 Hz). Projection filters are proposed to identify the state space model from closed-loop test data in time domain. A canonical transformation matrix is also derived to transform the identified state space model into the physical coordinate. The LAMSTF system is stabilized by using a linear quadratic regulator (LQR) feedback controller. The rate information is obtained by calculating the back difference of the sensed position signals. The reference inputs contain five uncorrelated random signals. This control input and the system reponse are recorded as input/output data to identify the system directly from the projection filters. The sampling time is 4 ms and the model is fairly accurate in predicting the step responses for different controllers while the analytical model has a deficiency in the pitch axis.

Huang, Jen-Kuang↗

Diagnostics and required R and D for control of DEMO grade plasmas

Even if the diagnostics of ITER performs as expected, installation and operation of the diagnostic systems in Demo device will be much harsher than those of the present ITER device. In order to operate the Demo grade plasmas, which may have a higher beta limit, safely with very limited number of simple diagnostic system, it requires a well defined predictable plasma modelling in conjunction with the reliable control system for burn control and potential harmful instabilities. Development of such modelling in ITER is too risky and the logical choice would be utilization of the present day steady state capable devices such as KSTAR and EAST. In order to fulfill this mission, sophisticated diagnostic systems such as 2D/3D imaging systems can validate the physics in the theoretical modeling and challenge the predictable capability.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Comparison of Observations and Predictions of Daytime Planetary-Boundary-Layer Heights and Surface Meteorological Variables in the Columbia River Gorge and Basin During the Second Wind Forecast Improvement Project

The second Wind Forecast Improvement Project (WFIP2) is an 18-month field campaign in the Pacific Northwest U.S.A., whose goal is to improve the accuracy of numerical-weather-prediction forecasts in complex terrain. The WFIP2 campaign involved the deployment of a large suite of in situ and remote sensing instrumentation, including eight 915-MHz wind-profiling radars, and surface meteorological stations. The evolution and annual variability of the daytime convective planetary-boundary-layer (PBL) height is investigated using the wind-profiling radars. Three models with different horizontal grid spacing are evaluated: the Rapid Refresh, the High-Resolution Rapid Refresh, and its nested version. The results are used to assess errors in the prediction of PBL height within the experimental and control versions of the models, with the experimental versions including changes and additions to the model parametrizations developed during the field campaign, and the control version using the parametrizations present in the National Oceanic and Atmospheric Administration/National Centers for Environmental Prediction operational version of the models at the start of the project. Results show that the high-resolution models outperform the low-resolution versions, the experimental versions perform better compared with the control versions, model PBL height estimations are more accurate on cloud-free days, and model estimates of the PBL height growth rate are more accurate than model estimates of the rate of decay. Finally, using surface sensors, we assess surface meteorological variables, finding improved surface irradiance and, to a lesser extent, improved 2-m temperature in the experimental version of the model.

54 ENVIRONMENTAL SCIENCES↗

Efficient data-driven models for prediction and optimization of geothermal power plant operations

Increasing the capacity of geothermal energy as a renewable resource calls for development and deployment of efficient control and optimization technologies for geothermal power plants. A data-driven prediction and optimization model is presented as a cost-effective and efficient alternative to physics-based approach. The model predicts power output and operational cost by propagating the influence of control and disturbance variables within an artificial neural network (ANN). Numerical experiments with simulated and field data from a real geothermal power plant are first used to demonstrate the prediction performance of the ANN model. The model is then adopted to maximize the net predicted power production by automatically adjusting the working fluid circulation rate. The optimization performance of the model in evaluated using a thermodynamic flowsheet simulation model. The workflow is applied to model and control the effect of ambient temperature on an air-cooled binary cycle power plant, which is complex and costly to perform using a physics-based predictive model. As a result, the performance of the method is demonstrated by applying it to both simulated and field datasets from a binary cycle geothermal power plant.

15 GEOTHERMAL ENERGY↗

Application of the Concepts of Load State, Load Space, and Output Space to Strain-Gage Balances

The concepts of load state, load space, and output space are used to describe elements of the development of a load prediction process for a six–component strain–gage balance. The use of these terms adds clarity during the selection of the independent and dependent variables for the development of the load prediction process. First, the control volume model of a strain–gage balance is reviewed as it helps better understand the role of loads, outputs, and state variables during the load prediction. Then, the description of the absolute load datum in both the load space and the output space is discussed. It is concluded that the use of the difference between a raw output and the natural zero of a balance gage has a key advantage for the definition of the output space. In that case, the origin of the output space will also describe the absolute load datum of the balance. Finally, details of the application of the concepts of load state, load space, and output space to two commonly used balance load prediction methods are discussed.

strain-gage balance↗

Determining the potential productivity of food crops in controlled environments

The quest to determine the maximum potential productivity of food crops is greatly benefitted by crop growth models. Many models have been developed to analyze and predict crop growth in the field, but it is difficult to predict biological responses to stress conditions. Crop growth models for the optimal environments of a Controlled Environment Life Support System (CELSS) can be highly predictive. This paper discusses the application of a crop growth model to CELSS; the model is used to evaluate factors limiting growth. The model separately evaluates the following four physiological processes: absorption of PPF by photosynthetic tissue, carbon fixation (photosynthesis), carbon use (respiration), and carbon partitioning (harvest index). These constituent processes determine potentially achievable productivity. An analysis of each process suggests that low harvest index is the factor most limiting to yield. PPF absorption by plant canopies and respiration efficiency are also of major importance. Research concerning productivity in a CELSS should emphasize: (1) the development of gas exchange techniques to continuously monitor plant growth rates and (2) environmental techniques to reduce plant height in communities.

Bugbee, Bruce↗

An analytical approach for predicting pilot induced oscillations

The optimal control model (OCM) of the human pilot is applied to the study of aircraft handling qualities. Attention is focused primarily on longitudinal tasks. The modeling technique differs from previous applications of the OCM in that considerable effort is expended in simplifying the pilot/vehicle analysis. After briefly reviewing the OCM, a technique for modeling the pilot controlling higher order systems is introduced. Following this, a simple criterion or determining the susceptability of an aircraft to pilot induced oscillations (PIO) is formulated. Finally, a model-based metric for pilot rating prediction is discussed. The resulting modeling procedure provides a relatively simple, yet unified approach to the study of a variety of handling qualities problems.

Hess, R. A.↗

Dynamic response for thermal control and measurement and fast radiation thermometry

A preliminary evaluation was made by ORNL of a two-color ratio pyrometer (TCRP) for temperature control in the Modular Electromagnetic Levitation (MEL) experiment. A discussion was presented by Eric Spjut at the 1987 NASA Non-Contact Temperature Measurement Workshop (NASA Conf. Publ. 2503, pp. 182-213) in which he described the non-linear characteristics of the time response of TCPs. Researchers replicated his model and results and note that the non-linear response behavior is minimized for small temperature steps at high temperatures. They then used the predicted response in a model for a proportional or integral feedback controller and predicted the control characteristics for heating and cooling a 5-mm diameter sphere of niobium at high (1500 to 2750 K) temperatures. The analysis shows that for a slow (25-ms) time response for a commercial RCRP, overshoots of several hundred kelvins will result from a 100-K decrease in the setpoint, and temperature tracking errors of 14 to 45 K will occur for control temperature ramps of 1000K/s. For a fast (greater than 0.1 ms) time response, the overshoot and ramp response errors are largely eliminated.

Shepard, R. L.↗

Some new control theoretic models for human operator display monitoring

Control theoretic techniques are applied to develop two new models for predicting human operator performance when monitoring an automatically controlled system. In one case it is assumed that the human monitors the instruments in order to rapidly detect failures. The second approach assumes that the instruments are sampled to best reconstruct the system status information. The relation of these models to existing prediction schemes, e.g., equal attention and the Sender's (1964) model is explored. It is concluded that a combination of failure detection and status estimation models offers the best potential for human operator application.

Kleinman, D. L.↗

Some new control theoretic models for human operator display monitoring

Control theoretic techniques are applied to develop two new models for predicting human operator performance when monitoring an automatically controlled system. In one case it is assumed that the human monitors the instruments in order to rapidly detect failures. A second approach assumes that the instruments are sampled to best reconstruct the system status information. The relation of these models to existing prediction schemes, e.g., equal attention and the Senders model is explored. It is concluded that a combination of failure detection and status estimation models offers the best potential for human operator application.

Kleinman, D. L.↗