SEARCH · Engineering Papers
Results for “prediction model”
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.
Nonlinear model predictive control and H robust control for a post-combustion CO2 capture process
Not Available
Determining the representative volume element size for three-dimensional microstructural material characterization. Part 1: Predictive models
Explore the source record for details and available documents.
Understanding the Impact of Unobservable Variables on the Performance of Predictive Models: The Need for Feature Space Partitioning and Fusion
When developing predictive models over a dataset, the model is globally optimized across the entire feature space to learn a decision boundary. However, when unobservable variables—which cannot be measured or estimated—interact with the observable variables, this can negatively impact the optimization applied to the decision boundary since the data samples introduced by unobservable variables may have little to no association with the applied global optimization. This, consequently, penalizes the entire decision boundary and model performance. This paper examines some of the detrimental effects of unobservable variables, particularly their role in creating new modes in the distribution of observable variables and reducing the separability of class distributions. Such challenges result in skewed or warped decision boundaries and decreased accuracy of model predictions, particularly for interpretable models like logistic regression and decision trees. Through two illustrative case examples, we highlight the need to address the challenges imposed by unobservable variables. We propose a strategy to mitigate these challenges by creating local regions within the feature space through partitioning. This enables the optimization of local models within the regions to overcome the impact of unobservability in different feature space localities. Research into a more sophisticated partitioning strategy and where the partition should be relative to the sample of interest is left as future work. Through the analysis of the impact of unobservability and the development of a partitioning method, we demonstrate the clear need for a partitioning strategy that integrates knowledge from multiple local models to estimate risk factors using information fusion. Thus, we establish the foundation and motivation for using partitioning and information fusion to overcome the effects of unobservability in predictive models. Formal fusion methods, such as Dempster-Shafer theory, can better leverage the information from local regions to improve the performance of interpretable predictive models in the presence of unobservable variables.
Machine learning based algorithms for uncertainty quantification in numerical weather prediction models
Complex numerical weather prediction models incorporate a variety of physical processes, each described by multiple alternative physical schemes with specific parameters. The selection of the physical schemes and the choice of the corresponding physical parameters during model configuration can significantly impact the accuracy of model forecasts. There is no combination of physical schemes that works best for all times, at all locations, and under all conditions. It is therefore of considerable interest to understand the interplay between the choice of physics and the accuracy of the resulting forecasts under different conditions. This paper demonstrates the use of machine learning techniques to study the uncertainty in numerical weather prediction models due to the interaction of multiple physical processes. The first problem addressed herein is the estimation of systematic model errors in output quantities of interest at future times, and the use of this information to improve the model forecasts. The second problem considered is the identification of those specific physical processes that contribute most to the forecast uncertainty in the quantity of interest under specified meteorological conditions. In order to address these questions we employ two machine learning approaches, random forests and artificial neural networks. The discrepancies between model results and observations at past times are used to learn the relationships between the choice of physical processes and the resulting forecast errors. Numerical experiments are carried out with the Weather Research and Forecasting (WRF) model. The output quantity of interest is the model precipitation, a variable that is both extremely important and very challenging to forecast. The physical processes under consideration include various micro-physics schemes, cumulus parameterizations, short wave, and long wave radiation schemes. The experiments demonstrate the strong potential of machine learning approaches to aid the study of model errors.
Estimates of Southern Hemispheric Gravity Wave Momentum Fluxes across Observations, Reanalyses, and Kilometer-Scale Numerical Weather Prediction Model
Abstract Gravity waves (GWs) are among the key drivers of the meridional overturning circulation in the mesosphere and upper stratosphere. Their representation in climate models suffers from insufficient resolution and limited observational constraints on their parameterizations. This obscures assessments of middle atmospheric circulation changes in a changing climate. This study presents a comprehensive analysis of stratospheric GW activity above and downstream of the Andes from 1 to 15 August 2019, with special focus on GW representation ranging from an unprecedented kilometer-scale global forecast model (1.4 km ECMWF IFS), ground-based Rayleigh lidar (CORAL) observations, modern reanalysis (ERA5), to a coarse-resolution climate model (EMAC). Resolved vertical flux of zonal GW momentum (GWMF) is found to be stronger by a factor of at least 2–2.5 in IFS compared to ERA5. Compared to resolved GWMF in IFS, parameterizations in ERA5 and EMAC continue to inaccurately generate excessive GWMF poleward of 60°S, yielding prominent differences between resolved and parameterized GWMFs. A like-to-like validation of GW profiles in IFS and ERA5 reveals similar wave structures. Still, even at ∼1 km resolution, the resolved waves in IFS are weaker than those observed by lidar. Further, GWMF estimates across datasets reveal that temperature-based proxies, based on midfrequency approximations for linear GWs, overestimate GWMF due to simplifications and uncertainties in GW wavelength estimation from data. Overall, the analysis provides GWMF benchmarks for parameterization validation and calls for three-dimensional GW parameterizations, better upper-boundary treatment, and vertical resolution increases commensurate with increases in horizontal resolution in models, for a more realistic GW analysis. Significance Statement Gravity wave–induced momentum forcing forms a key component of the middle atmospheric circulation. However, complete knowledge of gravity waves, their atmospheric effects, and their long-term trends are obscured due to limited global observations, and the inability of current climate models to fully resolve them. This study combines a kilometer-scale forecast model, modern reanalysis, and a coarse-resolution climate model to first compare the resolved and parameterized momentum fluxes by gravity waves generated over the Andes, and then evaluate the fluxes using a state-of-the-art ground-based Rayleigh lidar. Our analysis reveals shortcomings in current model parameterizations of gravity waves in the middle atmosphere and highlights the sensitivity of the estimated flux to the formulation used.
A Spatiotemporal Prediction Model for Regional Scheduling of Shared Bicycles Based on the INLA Method
Dock-less bicycle-sharing programs have been widely accepted as an efficient mode to benefit health and reduce congestions. And modeling and prediction has always been a core proposition in the field of transportation. Most of the existing demand prediction models for shared bikes take regions as research objects; therefore, a POI-based method can be a beneficial complement to existing research, including zone-level, OD-level, and station-level techniques. Point of interest (POI) is the location description of spatial entities, which can reflect the cycling route characteristics for both commuting and noncommuting trips to a certain extent, and is also the main generating point and attraction point of shared-bike travel flow. In this study, we make an effort to model a POI-level cycling demand with a Bayesian hierarchical method. The proposed model combines the integrated nested Laplace approximation (INLA) and random partial differential equation (SPDE) to cope with the huge computation in the modeling process. In particular, we have adopted the dock-less bicycle-sharing rental records of Mobike as a case study to validate our method; the study area was one of the fastest growing urban districts in Shanghai in August 2016. The operation results show that the method can help better understand, measure, and characterize spatiotemporal patterns of bike-share ridership at the POI level and quantify the impact of the spatiotemporal effect on bicycle-sharing use.
Station Keeping with an Autonomous Underwater Glider Using a Predictive Model of Ocean Currents
We investigate the use of an autonomous underwater glider as a platform for a virtual mooring. Our approach uses a simple vehicle motion model, a predictive model of ocean currents, and a greedy search algorithm in order to simulate possible actions available to the vehicle and select an action to minimize the distance from the target point. Results from a 19 day experiment in October 2016 near Monterey Bay are presented where we test our control algorithm as well as investigate the effect of a glider’s dive profile on its ability to act as a virtual mooring.
Real-time Nonlinear Model Predictive Control (NMPC) Strategies using Physics-Based Models for Advanced Lithium-ion Battery Management System (BMS)
Optimal operation of lithium-ion batteries requires robust battery models for advanced battery management systems (ABMS). A nonlinear model predictive control strategy is proposed that directly employs the pseudo-two-dimensional (P2D) model for making predictions. Using robust and efficient model simulation algorithms developed previously, the computational time of the nonlinear model predictive control algorithm is quantified, and the ability to use such models for nonlinear model predictive control for ABMS is established.
Erratum: “Reduced model prediction of electron temperature profiles in microtearing-dominated National Spherical Torus eXperiment plasmas” [Phys. Plasmas 21, 082510 (2014)]
No abstract prepared.
Sociomarkers and biomarkers: predictive modeling in identifying pediatric asthma patients at risk of hospital revisits
The importance of social components of health has been emphasized both in epidemiology and public health. This paper highlights the significant impact of social components on health outcomes in a novel way. Introducing the concept of sociomarkers, which are measurable indicators of social conditions in which a patient is embedded, we employed a machine learning approach that uses both biomarkers and sociomarkers to identify asthma patients at risk of a hospital revisit after an initial visit with an accuracy of 66%. The analysis has been performed over an integrated dataset consisting of individual-level patient information such as gender, race, insurance type, and age, along with ZIP code-level sociomarkers such as poverty level, blight prevalence, and housing quality. Using this uniquely integrated database, we then compare the traditional biomarker-based risk model and the sociomarker-based risk model. A biomarker-based predictive model yields an accuracy of 65% and the sociomarkerbased model predicts with an accuracy of 61%. Without knowing specific symptom-related features, the sociomarker-based model can correctly predict two out of three patients at risk. We systematically show that sociomarkers play an important role in predicting health outcomes at the individual level in pediatric asthma cases. Additionally, by merging multiple data sources with detailed neighborhood-level data, we directly measure the importance of residential conditions for predicting individual health outcomes.
Control of a nonlinear plant with two level input using a tabular adaptive-predictive model.
Tabular adaptive predictive model for trajectory following control of nonlinear plant with switched two-level input
Predictive modeling of pedestal structure in KSTAR using EPED model
A predictive calculation is given for the structure of edge pedestal in the H-mode plasma of the KSTAR (Korea Superconducting Tokamak Advanced Research) device using the EPED model. Particularly, the dependence of pedestal width and height on various plasma parameters is studied in detail. The two codes, ELITE and HELENA, are utilized for the stability analysis of the peeling-ballooning and kinetic ballooning modes, respectively. Summarizing the main results, the pedestal slope and height have a strong dependence on plasma current, rapidly increasing with it, while the pedestal width is almost independent of it. The plasma density or collisionality gives initially a mild stabilization, increasing the pedestal slope and height, but above some threshold value its effect turns to a destabilization, reducing the pedestal width and height. Among several plasma shape parameters, the triangularity gives the most dominant effect, rapidly increasing the pedestal width and height, while the effect of elongation and squareness appears to be relatively weak. Implication of these edge results, particularly in relation to the global plasma performance, is discussed.
Development of a Multicomponent Prediction Model for Acute Esophagitis in Lung Cancer Patients Receiving Chemoradiotherapy
Purpose: To construct a model for the prediction of acute esophagitis in lung cancer patients receiving chemoradiotherapy by combining clinical data, treatment parameters, and genotyping profile. Patients and Methods: Data were available for 273 lung cancer patients treated with curative chemoradiotherapy. Clinical data included gender, age, World Health Organization performance score, nicotine use, diabetes, chronic disease, tumor type, tumor stage, lymph node stage, tumor location, and medical center. Treatment parameters included chemotherapy, surgery, radiotherapy technique, tumor dose, mean fractionation size, mean and maximal esophageal dose, and overall treatment time. A total of 332 genetic polymorphisms were considered in 112 candidate genes. The predicting model was achieved by lasso logistic regression for predictor selection, followed by classic logistic regression for unbiased estimation of the coefficients. Performance of the model was expressed as the area under the curve of the receiver operating characteristic and as the false-negative rate in the optimal point on the receiver operating characteristic curve. Results: A total of 110 patients (40%) developed acute esophagitis Grade {>=}2 (Common Terminology Criteria for Adverse Events v3.0). The final model contained chemotherapy treatment, lymph node stage, mean esophageal dose, gender, overall treatment time, radiotherapy technique, rs2302535 (EGFR), rs16930129 (ENG), rs1131877 (TRAF3), and rs2230528 (ITGB2). The area under the curve was 0.87, and the false-negative rate was 16%. Conclusion: Prediction of acute esophagitis can be improved by combining clinical, treatment, and genetic factors. A multicomponent prediction model for acute esophagitis with a sensitivity of 84% was constructed with two clinical parameters, four treatment parameters, and four genetic polymorphisms.
Prediction modeling of physiological responses and human performance in the heat with application to space operations
This institute has developed a comprehensive USARIEM heat strain model for predicting physiological responses and soldier performance in the heat which has been programmed for use by hand-held calculators, personal computers, and incorporated into the development of a heat strain decision aid. This model deals directly with five major inputs: the clothing worn, the physical work intensity, the state of heat acclimation, the ambient environment (air temperature, relative humidity, wind speed, and solar load), and the accepted heat casualty level. In addition to predicting rectal temperature, heart rate, and sweat loss given the above inputs, our model predicts the expected physical work/rest cycle, the maximum safe physical work time, the estimated recovery time from maximal physical work, and the drinking water requirements associated with each of these situations. This model provides heat injury risk management guidance based on thermal strain predictions from the user specified environmental conditions, soldier characteristics, clothing worn, and the physical work intensity. If heat transfer values for space operations' clothing are known, NASA can use this prediction model to help avoid undue heat strain in astronauts during space flight.
Model Predictive Control-Based Dual-Mode Operation of an Energy-Stored Quasi-Z-Source Photovoltaic Power System
The energy-stored quasi-Z-source inverter (ES-qZSI) has attracted much attention for photovoltaic (PV) power generations, due to its capability to stabilize the PV power fluctuations with simple structure and other advantages of Z-source inverter. The improved dual-mode (DM) ES-qZSI is able to support all-weather operation even at night or cloudy days when PV power is extremely low. However, the traditional proportional–integral (PI) based control suffers from complicated controller design and poor dynamic response during mode transition, due to two sets of PI control required for the daytime and night operation modes. In order to overcome that, here in this article, we propose a model predictive control strategy for the DM-ES-qZSI PV power system. The system predictive models in both day and night operating modes are derived. The control strategy is disclosed to ensure high performance of the system, through calling the predictive models and defined cost functions of the two modes within a single control loop. Simulation and experiment are carried out to verify the effectiveness of the proposed control strategy.