Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “model evaluation”

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 1,729 records · Page 96

Evaluating Emulation-based Models of Distributed Computing Systems

Emulation-based models of distributed computing systems are collections of virtual ma- chines, virtual networks, and other emulation components configured to stand in for oper- ational systems when performing experimental science, training, analysis of design alterna- tives, test and evaluation, or idea generation. As with any tool, we should carefully evaluate whether our uses of emulation-based models are appropriate and justified. Otherwise, we run the risk of using a model incorrectly and creating meaningless results. The variety of uses of emulation-based models each have their own goals and deserve thoughtful evaluation. In this paper, we enumerate some of these uses and describe approaches that one can take to build an evidence-based case that a use of an emulation-based model is credible. Predictive uses of emulation-based models, where we expect a model to tell us something true about the real world, set the bar especially high and the principal evaluation method, called validation , is comensurately rigorous. We spend the majority of our time describing and demonstrating the validation of a simple predictive model using a well-established methodology inherited from decades of development in the compuational science and engineering community.

97 MATHEMATICS AND COMPUTING↗

Beam Hardening Artifact Reduction in X-Ray CT Reconstruction of 3D Printed Metal Parts Leveraging Deep Learning and CAD Models

Nondestructive evaluation (NDE) of additively manufactured (AM) parts is important for understanding the impacts of various process parameters and qualifying the built part. X-ray computed tomography (XCT) has played a critical role in rapid NDE and characterization of AM parts. However, XCT of metal AM parts can be challenging because of artifacts produced by standard reconstruction algorithms as a result of a confounding effect called “beam hardening.” Beam hardening artifacts complicate the analysis of XCT images and adversely impact the process of detecting defects, such as pores and cracks, which is key to ensuring the quality of the parts being printed. In this work, we propose a novel framework based on using available computer-aided design (CAD) models for parts to be manufactured, accurate XCT simulations, and a deep-neural network to produce high-quality XCT reconstructions from data that are affected by noise and beam hardening. Using extensive experiments with simulated data sets, we demonstrate that our method can significantly improve the reconstruction quality, thereby enabling better detection of defects compared with the state of the art. We also present promising preliminary results of applying the deep networks trained using CAD models to experimental data obtained from XCT of an AM jet-engine turbine blade.

Ziabari, Amir↗

Evaluation of Industry Standard Turbulence Models on an Axisymmetric Supersonic Compression Corner

Reynolds-averaged Navier-Stokes computations of a shock-wave/boundary-layer interaction (SWBLI) created by a Mach 2.85 flow over an axisymmetric 30-degree compression corner were carried out. The objectives were to evaluate four turbulence models commonly used in industry, for SWBLIs, and to evaluate the suitability of this test case for use in further turbulence model benchmarking. The Spalart-Allmaras model, Menter's Baseline and Shear Stress Transport models, and a low-Reynolds number k- model were evaluated. Results indicate that the models do not accurately predict the separation location; with the SST model predicting the separation onset too early and the other models predicting the onset too late. Overall the Spalart-Allmaras model did the best job in matching the experimental data. However there is significant room for improvement, most notably in the prediction of the turbulent shear stress. Density data showed that the simulations did not accurately predict the thermal boundary layer upstream of the SWBLI. The effect of turbulent Prandtl number and wall temperature were studied in an attempt to improve this prediction and understand their effects on the interaction. The data showed that both parameters can significantly affect the separation size and location, but did not improve the agreement with the experiment. This case proved challenging to compute and should provide a good test for future turbulence modeling work.

Turbulent Boundary Layer↗

Improving the energy response of external beam therapy (EBT) GafChromic{sup TM} dosimetry films at low energies (≤100 keV)

Purpose: Purpose of this work is to investigate the effects of varying the active layer composition of external beam therapy (EBT) GafChromic{sup TM} films on the energy dependence of the film, as well as try to develop a new prototype with more uniform energy response at low photon energies (⩽100 keV). Methods: First, the overall energy response (S{sub AD,} {sub W}(Q)) of different commercial EBT type film models that represent the three different generations produced to date, i.e., EBT, EBT2, and EBT3, was investigated. Pieces of each film model were irradiated to a fixed dose of 2 Gy to water for a wide range of beam qualities and the corresponding S{sub AD,} {sub W}(Q) was measured using a flatbed document scanner. Furthermore, the DOSRZnrc Monte Carlo code was used to determine the absorbed dose to water energy dependence of the film, f(Q). Moreover, the intrinsic energy dependence, k{sub bq}(Q), for each film model was evaluated using the corresponding S{sub AD,} {sub W}(Q) and f(Q). In the second part of this study, the authors investigated the effects of changing the chemical composition of the active layer on S{sub AD,} {sub W}(Q). Finally, based on these results, the film manufacturer fabricated several film prototypes and the authors evaluated their S{sub AD,} {sub W}(Q). Results: The commercial EBT film model shows an under response at all energies below 100 keV reaching 39% ± 4% at about 20 keV. The commercial EBT2 and EBT3 film models show an under response of about 27% ± 4% at 20 keV and an over response of about 16% ± 4% at 40 keV.S{sub AD,} {sub W}(Q) of the three commercial film models at low energies show strong correlation with the corresponding f{sup −1}(Q) curves. The commercial EBT3 model with 4% Cl in the active layer shows under response of 22% ± 4% at 20 keV and 6% ± 4% at about 40 keV. However, increasing the mass percent of chlorine makes the film more hygroscopic which may affect the stability of the film's readout. The EBT3 film prototype with 7.5% Si shows a significant improvement in the energy response at very low energies compared to the commercial EBT3 films with 4% Cl. It shows under response of 15% ± 5% at about 20 keV to 2% ± 5% at about 40 keV. However, according to the manufacturer, the addition of 7.5% Si as SiO{sub 2} adversely affected the viscosity of the active fluid and therefore affected the potential use in commercial machine coating. The latest commercial EBT3 film model with 7% Al as Al{sub 2}O{sub 3} shows an overall improvement in S{sub AD,} {sub W}(Q) compared to previous commercial EBT3 films. It shows under response at all energies <100 keV, varying from 20% ± 4% at 20 keV to 6% ± 4% at 40 keV. Conclusions: The energy response of films in the energy range <100 keV can be improved by adjusting the active layer chemical composition. Removing bromine eliminated the over response at about 40 keV. The under response at energies ≤30 keV is improved by adding 7% Al to the active layer in the latest commercial EBT3 film models.

61 RADIATION PROTECTION AND DOSIMETRY↗

A New Regional Climate Model for POLAR‐CORDEX: Evaluation of a 30‐Year Hindcast with COSMO‐CLM^2 Over Antarctica

Continent‐wide climate information over the Antarctic Ice Sheet (AIS) is important to obtain accurate information of present climate and reduce uncertainties of the ice sheet mass balance response and resulting global sea level rise to future climate change. In this study, the COSMO‐CLM^2 Regional Climate Model is applied over the AIS and adapted for the specific meteorological and climatological conditions of the region. A 30‐year hindcast was performed and evaluated against observational records consisting of long‐term ground‐based meteorological observations, automatic weather stations, radiosoundings, satellite records, stake measurements and ice cores. Reasonable agreement regarding the surface and upper‐air climate is achieved by the COSMO‐CLM^2 model, comparable to the performance of other state‐of‐the‐art climate models over the AIS. Meteorological variability of the surface climate is adequately simulated, and biases in the radiation and surface mass balance are small. The presented model therefore contributes as a new member to the COordinated Regional Downscaling EXperiment project over the AIS (POLAR‐CORDEX) and the CORDEX‐CORE initiative.

Antarctica↗

Evaluation of a Model-Based Groundwater Drought Indicator in the Conterminous U.S.

Monitoring groundwater drought using land surface models is a valuable alternative given the current lack of systematic in situ measurements at continental and global scales and the low resolution of current remote sensing based groundwater data. However, uncertainties inherent to land surface models may impede drought detection, and thus should be assessed using independent data sources. In this study, we evaluated a groundwater drought index (GWI) derived from monthly groundwater storage output from the Catchment Land Surface Model (CLSM) using a GWI similarly derived from in situ groundwater observations. Groundwater observations were obtained from unconfined or semi-confined aquifers in eight regions of the central and northeastern U.S. Regional average GWI derived from CLSM exhibited strong correlation with that from observation wells, with correlation coefficients between 0.43 and 0.92. GWI from both in situ data and CLSM was generally better correlated with the Standard Precipitation Index (SPI) at 12 and 24 month timescales than at shorter timescales, but it varied depending on climate conditions. The correlation between CLSM derived GWI and SPI generally decreases with increasing depth to the water table, which in turn depends on both bedrock depth (a CLSM parameter) and mean annual precipitation. The persistence of CLSM derived GWI is spatially varied and again shows a strong influence of depth to groundwater. CLSM derived GWI generally persists longer than GWI derived from in situ data, due at least in part to the inability of coarse model inputs to capture high frequency meteorological variability at local scales. The study also showed that groundwater can have a significant impact on soil moisture persistence where the water table is shallow. Soil moisture persistence was estimated to be longer in the eastern U.S. than in the west, in contrast to previous findings that were based on models that did not represent groundwater. Assimilation of terrestrial water storage data from the Gravity Recovery and Climate Experiment (GRACE) satellite mission improved the correlation between CLSM based regional average GWI and that based on in situ data in six of the eight regions. Practical issues regarding the application of GRACE assimilated groundwater storage for drought detection are discussed. An important conclusion of this study is that model parameters that control the depth to the water table, including bedrock depth, strongly influence the evolution and persistence of simulated groundwater and require careful configuration for drought monitoring.

Evaluation↗

Evaluation of advanced geopotential models for operational orbit determination

To meet future orbit determination accuracy requirements for different NASA projects, analyses are performed using Tracking and Data Relay Satellite System (TDRSS) tracking measurements and orbit determination improvements in areas such as the modeling of the Earth's gravitational field. Current operational requirements are satisfied using the Goddard Earth Model-9 (GEM-9) geopotential model with the harmonic expansion truncated at order and degree 21 (21-by-21). This study evaluates the performance of 36-by-36 geopotential models, such as the GEM-10B and Preliminary Goddard Solution-3117 (PGS-3117) models. The Earth Radiation Budget Satellite (ERBS) and LANDSAT-5 are the spacecraft considered in this study.

Radomski, M. S.↗

Effects of Mineral Surface Properties on Supercritical CO 2 Wettability in a Siliciclastic Reservoir

Wettability is a key reservoir characteristic influencing geological carbon sequestration (GCS) processes, such as CO 2 transport and storage capacity. Wettability is often determined on a limited number of reservoir samples by measuring the contact angle at the CO 2 /brine/mineral interface, but the ability to predict this value remains a challenge. In this work, minerals comprising a natural reservoir sample were identified, and the influence of their surface roughness and mineralogy on the contact angle was quantified to evaluate predictive models and controlling mechanisms. The natural sample was obtained from the Mount Simon formation, a representative siliciclastic reservoir that is the site of a United States Department of Energy CO 2 injection project. A thin section of the Mount Simon sandstone was examined with compound light microscopy and environmental scanning electron microscopy (ESEM) coupled with energy-dispersive X-ray spectroscopy (EDS). Quartz and feldspar were identified as dominant minerals and were coated with various reddish black precipitates consistent with illite clay and iron oxide hematite. Contact angle (θ) measurements were conducted for the four representative minerals and the Mount Simon sample over a range of pressures (2–25 MPa) at 40 °C. At supercritical conditions, all samples are strongly water-wet, with contact angles between 27° and 45°. Several predictive models for contact angle were evaluated for the mineral and Mount Simon samples, including the Wenzel and Cassie–Baxter models, plus newly proposed modifications of these that account for the fraction of different minerals comprising the reservoir sample surface, the surface roughness, and the extent that roughness pits are filled with brine. Modeling results suggest that the fraction of mineral surfaces containing roughness pits filled with brine is the most important reservoir characteristic that controls wettability in the Mount Simon sandstone, followed by surface mineralogy. Finally, to our knowledge, this is one of the few studies to investigate the effects of individual minerals on the wettability of a natural reservoir sample.

58 GEOSCIENCES↗

Modeling prismatic HTGRs with U.S. N.R.C advanced gas reactor evaluator (AGREE)

A core fluids and heat transfer model has been developed for the prismatic high temperature gas reactor in support of the US NRC Next Generation Nuclear Plant (NGNP) evaluation model. The core fluids modeling relies on a subchannel approach in which the primary coolant flow path through the core region and vertical in-core and ex-core gaps can be modeled as individual subchannels. These subchannels are connected together to represent a three dimensional reactor. An initial validation calculation for the core fluids model has been performed using data available in literature for bypass flow. The predicted bypass flow was within 2.6% of the value reported in the literature. The core level heat transfer model is based on a triangular finite volume method, where the base triangle is one sixth of the prismatic block. In order to improve the spatial accuracy at this level, a triangular refinement method was also implemented. The fuel compact temperature is calculated by a cylindrical conduction model which is implicitly coupled to the triangular core level model. The preliminary verification of the model was performed by comparing AGREE to a finite element code COMSOL by analyzing the MHTGR core heat transfer. Further verification and validation is currently an ongoing effort. (authors)

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

Evaluation of Immunocompetent Mouse Models for Borrelia miyamotoi Infection

Borrelia miyamotoi is a relapsing fever spirochete that is harbored by Ixodes spp. ticks and is virtually uncharacterized, compared to other relapsing fever Borrelia vectored by Ornithodoros spp. ticks. There is not an immunocompetent mouse model for studying B. miyamotoi infection in vivo or for transmission in the vector-host cycle. Our goal was to evaluate B. miyamotoi infections in multiple mouse breeds/strains as a prelude to the ascertainment of the best experimental infection model. Two B. miyamotoi strains, namely, LB-2001 and CT13-2396, as well as three mouse models, namely, CD-1, C3H/HeJ, and BALB/c, were evaluated. We were unable to observe B. miyamotoi LB-2001 spirochetes in the blood via darkfield microscopy or to detect DNA via real-time PCR post needle inoculation in the CD-1 and C3H/HeJ mice. However, LB-2001 DNA was detected via real-time PCR in the blood of the BALB/c mice after needle inoculation, although spirochetes were not observed via microscopy. CD-1, C3H/HeJ, and BALB/c mice generated an antibody response to B. miyamotoi LB-2001 following needle inoculation, but established infections were not detected, and the I. scapularis larvae failed to acquire spirochetes from the exposed CD-1 mice. In contrast, B. miyamotoi CT13-2396 was visualized in the blood of the CD-1 and C3H/HeJ mice via darkfield microscopy and detected by real-time PCR post needle inoculation. Both mouse strains seroconverted. However, no established infection was detected in the mouse organs, and the I. scapularis larvae failed to acquire Borrelia after feeding on CT13-2396 exposed CD-1 or C3H/HeJ mice. Furthermore, these findings underscore the challenges in establishing an experimental B. miyamotoi infection model in immunocompetent laboratory mice.

60 APPLIED LIFE SCIENCES↗

Bridging Theory with Practice: An Exploratory Study of Visualization Use and Design for Climate Model Comparison

Evaluation methodologies in visualization have mostly focused on how well the tools and techniques cater to the analytical needs of the user. While this is important in determining the effectiveness of the tools and advancing the state-of-the-art in visualization research, a key area that has mostly been overlooked is how well established visualization theories and principles are instantiated in practice. This is especially relevant when domain experts, and not visualization researchers, design visualizations for analysis of their data or for broader dissemination of scientific knowledge. There is very little research on exploring the synergistic capabilities of cross-domain collaboration between domain experts and visualization researchers. To fill this gap, in this paper we describe the results of an exploratory study of climate data visualizations conducted in tight collaboration with a pool of climate scientists. The study analyzes a large set of static climate data visualizations for identifying their shortcomings in terms of visualization design. The outcome of the study is a classification scheme that categorizes the design problems in the form of a descriptive taxonomy. The taxonomy is a first attempt for systematically categorizing the types, causes, and consequences of design problems in visualizations created by domain experts. We demonstrate the use of the taxonomy for a number of purposes, such as, improving the existing climate data visualizations, reflecting on the impact of the problems for enabling domain experts in designing better visualizations, and also learning about the gaps and opportunities for future visualization research. We demonstrate the applicability of our taxonomy through a number of examples and discuss the lessons learnt and implications of our findings.

54 ENVIRONMENTAL SCIENCES↗

Evaluating the Nation's Pipeline Infrastructure with NETL's Advanced Infrastructure Integrity Model (AIIM)

This poster is a part of BIL-EDX4CCS Task 36: Advanced Infrastructure Integrity Modeling to Evaluate Existing Energy Infrastructure Reusability and Risk, the goal of which is to produce a smart tool that will assess existing energy infrastructure reusability and risk using the Advanced Infrastructure Integrity Model (AIIM). This model forecasts lifespan and potential risk using a multitude of factors such as incidents reports, structural characteristics, and the surrounding environment. The project aims to provide scientific insights for a better understanding of carbon storage (CS), potential to support CS stakeholder needs, national decarbonization, and mitigating climate change. AIIM will utilize an energy infrastructure database as its input, developed by acquiring publicly available data as well as NETL derived products. These resources include incidents, geohazards, and infrastructure variables. Soil data in the form of rasters and pipeline incident reports were processed and a script was developed to count the number of times features such as roads, railroads, and rivers intersected with pipeline segments which were then converted to points. Distance to oil and natural gas wells, petroleum ports, intermodal freight facilities, and geologic structures were also calculated. After data preparation and quality control was completed, the data was integrated into the pipeline points. Once models are complete, a smart tool will be created in the form of an online dashboard.

Malay, Caleb↗

Estimation of Eddy Dissipation Rates from Mesoscale Model Simulations

The Eddy Dissipation Rate is an important metric for representing the intensity of atmospheric turbulence and is used as an input parameter for predicting the decay of aircraft wake vortices. In this study, the forecasts of eddy dissipation rates obtained from the current state-of-the-art mesoscale model are evaluated for terminal area applications. The Weather Research and Forecast mesoscale model is used to simulate the planetary boundary layer at high horizontal and vertical mesh resolutions. The Bougeault-Lacarrer and the Mellor-Yamada-Janji schemes implemented in the Weather Research and Forecast model are evaluated against data collected during the National Aeronautics and Space Administration s Memphis Wake Vortex Field Experiment. Comparisons with other observations are included as well.

Ahmad, Nashat N.↗

Evaluating the Community Land Model in a pine stand with shading manipulations and 13 CO 2 labeling

Carbon partitioning and flow through ecosystems regulates land surface atmosphere CO 2 exchange and thus is a key, albeit uncertain component of mechanistic models. The Partitioning in Trees and Soil (PiTS) experiment-model project tracked C partitioning through a young Pinus taeda stand following pulse-labeling with 13 CO 2 and two levels of shading. The field component of this project provided process-oriented data that was used to evaluate and improve terrestrial biosphere model simulations of rapid shifts in carbon partitioning and hydrological dynamics under varying environmental conditions. Here we tested the performance of the Community Land Model version 4 (CLM4) in capturing short-term carbon and water dynamics in relation to manipulative shading treatments, and the timing and magnitude of carbon fluxes through various compartments of the ecosystem. To constrain CLM4 to closely simulate pretreatment conditions, we calibrated select model parameters with the pretreatment observational data. Compared to CLM4 simulations with default parameters, CLM4 with calibrated model parameters was better able to simulate pretreatment vegetation carbon pools, light response curves, and other initial states and fluxes of carbon and water. Over a 3-week treatment period, the calibrated CLM4 generally reproduced the impacts of shading on average soil moisture at 15-95 cm depth, transpiration, relative change in stem carbon, and soil CO 2 efflux rate, although some discrepancies in the estimation of magnitudes and temporal evolutions existed. CLM4, however, was not able to track the progression of the 13 CO 2 label from the atmosphere through foliage, phloem, roots or surface soil CO 2 efflux, even when optimized model parameters were used. This model bias arises, in part, from the lack of a short-term non-structural carbohydrate storage pool and progressive timing of within-plant transport, thus indicating a need for future work to improve the allocation routines in CLM4. Overall, these types of detailed evaluations of CLM4, paired with intensive field manipulations, can help to identify model strengths and weaknesses, model uncertainties, and additional observations necessary for future model development.

54 ENVIRONMENTAL SCIENCES↗

A critical evaluation of various turbulence models as applied to internal fluid flows

Models employed in the computation of turbulent flows are described and their application to internal flows is evaluated by examining the predictions of various turbulence models in selected flow configurations. The main conclusions are: (1) the k-epsilon model is used in a majority of all the two-dimensional flow calculations reported in the literature; (2) modified forms of the k-epsilon model improve the performance for flows with streamline curvature and heat transfer; (3) for flows with swirl, the k-epsilon model performs rather poorly; the algebraic stress model performs better in this case; and (4) for flows with regions of secondary flow (noncircular duct flows), the algebraic stress model performs fairly well for fully developed flow, for developing flow, the algebraic stress model performance is not good; a Reynolds stress model should be used. False diffusion and inlet boundary conditions are discussed. Countergradient transport and its implications in turbulence modeling is mentioned. Two examples of recirculating flow predictions obtained using PHOENICS code are discussed. The vortex method, large eddy simulation (modeling of subgrid scale Reynolds stresses), and direct simulation, are considered. Some recommendations for improving the model performance are made. The need for detailed experimental data in flows with strong curvature is emphasized.

Nallasamy, M.↗

Porous Gelatin Particles for Uterine Artery Embolization: An Experimental Study of Intra-Arterial Distribution, Uterine Necrosis, and Inflammation in a Porcine Model

PurposeWe evaluated the location of porous gelatin particles (GP; Gelpart; Nippon Kayaku/Astellas, Tokyo, Japan) within the arterial vasculature and their acute effects on uterine necrosis and inflammation after uterine artery embolization (UAE) in swine.Materials and MethodsAdult nonpregnant pigs (n = 6) were allocated to either 1- (n = 3) or 2-mm GP (n = 3). Superselective and bilateral embolization of the uterine arteries was performed. All animals were killed 1 week after UAE. Macroscopic and microscopic findings, including the level of arterial occlusion and their effect on uterine necrosis and inflammation, were analyzed.ResultsAll UAE procedures were completed without severe complications. The macroscopic necrosis was seen in two animals in the 2-mm group with an extent of <50%. The location of the occluded arteries did not differ significantly between groups. The median diameters of the occluded arteries were 449 {mu}m (95% confidence interval [CI] 417-538 {mu}m) in the 1-mm GP group and 484 {mu}m (95% CI 370-560 {mu}m) in the 2-mm GP group. As for microscopic necrosis, no statistically significant difference was observed. The qualitative inflammatory reaction was significantly greater in the 2-mm GP group than in the 1-mm group (p < 0.001).ConclusionsBoth 1- and 2-mm GP occluded the arteries relevant to the target diameter for UAE in porcine uterus, presumably due to the plastic deformity. Both sizes of GP were associated with limited areas of necrosis; however, evaluation of inflammatory reaction was preliminary. Further study with adequate evaluation of inflammatory reactions is suggested.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Bariatric Arterial Embolization with Non-spherical Polyvinyl Alcohol Particles for Ghrelin Suppression in a Swine Model

PurposeTo evaluate the effect of bariatric arterial embolization (BAE) with non-spherical polyvinyl alcohol (PVA) particles on systemic ghrelin levels, weight change, and gastric ulceration risk in a swine model.Materials and MethodsFrom March 2014 to February 2015, ten healthy swine were used in the study (mean weight 31.5 kg; range 24.0–41.5 kg). The animals were randomly assigned to two groups: the embolized group (n = 5) in which BAE was performed and the control group (n = 5). In the embolized group, BAE was performed by selectively infusing 150–250 or 50–150 μm PVA into the gastric arteries that supplied the fundus of the stomach. In the control group, a sham procedure was performed with saline infusion. Plasma ghrelin levels were prospectively obtained at baseline and every 2 weeks thereafter. Endoscopy was performed 3 weeks after BAE to see whether any gastric ulcer occurred. To determine the durability of the occluded arteries, repeated celiac trunk angiography was performed 8 weeks after BAE. Then, all the swine were killed and necropsies were performed.ResultsThe mean post-procedure ghrelin value decreased by 370.0 pg/mL in the embolized group at week 3 (mean 536.0 ± 334.3 pg/mL) and week 5 (mean 515.0 ± 150.0 pg/mL, p < 0.05) relative to baseline (880.0 ± 559.5 pg/mL), respectively, but ghrelin levels were not significantly decreased between the embolized and control groups. There was a significant body weight change as follows: 35.1 ± 9.5 to 46.6 ± 15.7 kg and 31.8 ± 5.8 to 41.2 ± 6.6 kg at baseline and endpoint in the control and embolized groups, respectively (p < 0.05). However, the difference between groups was not significant at endpoint. In the embolized group, ulcerations were identified in three animals (60%) and the recanalization of the embolized arteries was noted on follow-up angiography in three animals (60%), respectively.ConclusionBAE with PVA particles can transiently suppress ghrelin levels in embolized animals, but levels are not significantly different compared to controls. Stomach ulcerations were identified in 60% of the embolized animals.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

A novel improved model for building energy consumption prediction based on model integration

Building energy consumption prediction plays an irreplaceable role in energy planning, management, and conservation. Constantly improving the performance of prediction models is the key to ensuring the efficient operation of energy systems. Moreover, accuracy is no longer the only factor in revealing model performance, it is more important to evaluate the model from multiple perspectives, considering the characteristics of engineering applications. Based on the idea of model integration, this paper proposes a novel improved integration model (stacking model) that can be used to forecast building energy consumption. The stacking model combines advantages of various base prediction algorithms and forms them into “meta-features” to ensure that the final model can observe datasets from different spatial and structural angles. Two cases are used to demonstrate practical engineering applications of the stacking model. A comparative analysis is performed to evaluate the prediction performance of the stacking model in contrast with existing well-known prediction models including Random Forest, Gradient Boosted Decision Tree, Extreme Gradient Boosting, Support Vector Machine, and K-Nearest Neighbor. The results indicate that the stacking method achieves better performance than other models, regarding accuracy (improvement of 9.5%–31.6% for Case A and 16.2%–49.4% for Case B), generalization (improvement of 6.7%–29.5% for Case A and 7.1%-34.6% for Case B), and robustness (improvement of 1.5%–34.1% for Case A and 1.8%–19.3% for Case B). The proposed model enriches the diversity of algorithm libraries of empirical models.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗