Search NASA⌕ Search

SEARCH · Search NASA

Results for “Building performance simulation”

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 1,099 records · Page 61

A new biogeochemical modelling framework (FLaMe-v1.0) for lake methane emissions on the regional scale: development and application to the European domain

This study presents a new physical-biogeochemical modelling framework for simulating lake methane (CH 4 ) emissions at regional scales. The new model, FLaMe-v1.0 (Fluxes of Lake Methane), rests on an innovative, computationally efficient lake clustering approach that enables the simulation of CH 4 emissions across a large number of lakes. Building on the Canadian Small Lake Model (CSLM) that simulates the lake physics, we develop a suite of biogeochemical modules to simulate transient dynamics of organic Carbon (C), Oxygen (O 2 ), and CH 4 . We first test the performance of FLaMe-v1.0 by analyzing physical and biogeochemical processes in two theoretical lakes with characteristics that can be considered representative for many lakes (an oligotrophic, deep lake driven by cold climate versus a eutrophic, shallow lake driven by warm climate). Next, we evaluate the model by comparing simulated and observed timeseries of CH 4 emissions in four well-surveyed lakes. We then apply FLaMe-v1.0 at the European scale to evaluate simulated diffusive and ebullitive lake CH 4 fluxes against in-situ measurements in both boreal and central European regions. Finally, we provide a first assessment of the spatio-temporal variability in CH 4 emissions from European lakes with a surface area comprised between 0.1–1000 km 2 (n= 108 407, total area = 1.33 × 105 km 2 ), indicating a total emission of 0.97 ± 0.23 Tg CH 4 yr −1 , with the uncertainty constrained by combining FLaMe-v1.0 and machine learning techniques. Moreover, 30 % and 70 % of these CH 4 emissions are through diffusive and ebullitive pathways, respectively. Annually averaged CH 4 emission rates per unit lake area during 2010–2016 have a South-to-North decreasing gradient, resulting in a mean over the European domain as 7.39 g CH 4 m −2 yr −1 . Our simulations reveal a strong seasonality (with ice-blocking effects accounted for) in European lake CH 4 emissions, with nearly ten times higher emissions during late summer than during winter. This pronounced seasonal variation highlights the importance of accounting for the sub-annual variability in CH 4 emissions to accurately constrain regional CH 4 budgets. In the future, FLaMe-v1.0 could be embedded into Earth System Models to investigate the feedback between climate warming and global lake CH 4 emissions.

Maisonnier, Manon [Free Univ. of Brussels (Belgium↗

Software For Development Of Expert Systems

Modular software system helps create efficient artificial-intelligence computer programs. STAR*TOOL system: Set of high-level software tools; assists programmers in creation of efficient knowledge-based software systems. Provides language and capabilities for compilation of application programs written in Common LISP. Features modularity enabling elimination of unnecessary capabilities from final application program and achieves greater computing performance. Runs on any computer that supports Common LISP and has sufficient memory. Provides programmer with necessary software tools to build wide variety of reasoning and inference engines for such applications as planning, diagnosis and analysis, and simulation.

James, Mark L.↗

Coupled Stochastic Time-Inverted Lagrangian Transport/Weather Forecast and Research/Vegetation Photosynthesis and Respiration Model: Simulations of Tower-Based and Airborne CO2 Measurements - Part II

This paper evaluates simulations of atmospheric CO2 measured in 2004 at continental surface and airborne receptors, intended to test the capability to use data with high temporal and spatial resolution for analyses of carbon sources and sinks at regional and continental scales. The simulations were performed using the Stochastic Time-Inverted Lagrangian Transport (STILT) model driven by the Weather Forecast and Research (WRF) model, and linked to surface fluxes from the satellite-driven Vegetation Photosynthesis and Respiration Model (VPRM). The simulations provide detailed representations of hourly CO2 tower data and reproduce the shapes of airborne vertical profiles with high fidelity. WRF meteorology gives superior model performance compared with standard meteorological products, and the impact of including WRF convective mass fluxes in the STILT trajectory calculations is significant in individual cases. Important biases in the simulation are associated with the nighttime CO2 build-up and subsequent morning transition to convective conditions, and with errors in the advected lateral boundary condition. Comparison of STILT simulations driven by the WRF model against those driven by the Brazilian variant of the Regional Atmospheric Modeling System (BRAMS) shows that model-to-model differences are smaller than between an individual transport model and observations, pointing to systematic errors in the simulated transport. Future developments in the WRF model s data assimilation capabilities, basic research into the fundamental aspects of trajectory calculations, and intercomparison studies involving other transport models, are possible venues for reducing these errors. Overall, the STILT/WRF/VPRM offers a powerful tool for continental and regional scale carbon flux estimates.

Eluszkiewicz, Janusz↗

NASA’s Space Launch System: Progress Toward Launch

The Space Launch System (SLS), NASA’s cornerstone launch capability for a new generation of deep space exploration, has begun assembly at Kennedy Space Center (KSC) in preparation for launch in 2021. SLS will provide an unparalleled launch capability for human and robotic deep space exploration missions. Its proven propulsion system, upgrade path to more powerful vehicles, and high-volume payload fairings make it the foundation for ambitious and demanding as part of the Artemis program. Artemis is NASA’s 21st-century plan to put boots on the Moon and to perform sustainable, long-term science in deep space, with eyes toward sending explorers to Mars. The initial SLS vehicle to fly, Block 1 in the crew configuration – with the new Orion spacecraft – is scheduled to lift off from revitalized launch facilities at KSC in 2021 for an uncrewed test flight known as Artemis I. Manufacturing is complete on the vehicle and all elements have been delivered to the Exploration Ground Systems (EGS) Program, except the core stage. The program’s all-new development, the core stage is currently in the midst of a “Green Run” test campaign at Stennis Space Center (SSC). Eight progressively more challenging tests in the Green Run series will culminate in a hot-fire of four flight-proven liquid hydrogen/liquid oxygen (LH2/LOX) RS25 engines. Following the Green Run hotfiring, the core stage will ship to KSC. Already at KSC, aft sections of the fivesegment solid rocket boosters are being assembled. Previously delivered elements, including the upper stage, are undergoing final checkouts in preparation for stacking. An exciting 2021 will include such milestones as stacking SLS and Orion in KSC’s Vehicle Assembly Building (VAB), modal testing, roll out to Launch Pad 39B, Wet Dress Rehearsal (WDR), and launch. Teams across the country are preparing for launch by finalizing procedures, defining launch constraints and flight rules, training console operators, performing simulations, and more. With the SLS Block 1 vehicle for Artemis I nearing integration and launch, the second Block 1 vehicle in the crew configuration, which will carry astronauts on an Artemis II hybrid lunar flyby mission, has several elements manufactured. In fact, the solid rocket motor segments and RS-25 engines are complete. Those program elements are processing hardware for the third flight and working toward manufacturing the second SLS variant to fly, Block 1B, which will onramp a powerful new upper stage, the Exploration Upper Stage (EUS).

John Honeycutt↗

Analysis of Radiant Cooling System Configurations Integrated with Cooling Tower for Different Indian Climatic Zones

Radiant cooling system has proven to be a low energy consumption system for building cooling needs. This study describes the use of cooling tower in radiant cooling system to improve the overall system efficiency. A comprehensive simulation feasibility study of the application of cooling tower in radiant cooling system was performed for the fifteen cities in different climatic zones of India. It was found that in summer, the wet bulb temperature (WBT) of the different climatic zones except warm-humid is suitable for the integration of cooling tower with radiant cooling system. In these climates, cooling tower can provide on average 24 C to 27 C water In order to achieve the energy saving potential, three different configurations of radiant cooling system have been compared in terms of energy consumption. The different configurations of the radiant cooling system integrated with cooling tower are: (1) provide chilled water to the floor, wall and ceiling mounted tubular installation. (2) provide chilled water to the wall and ceiling mounted tabular installation. In this arrangement a separate chiller has also been used to provide chilled water at 16 C to the floor mounted tubular installation. (3) provide chilled water to the wall mounted tabular installation and a separate chiller is used to provide chilled water at 16 C to the floor and ceiling mounted tabular installation. A dedicated outdoor air system is also coupled for dehumidification and ventilation in all three configurations. A conventional all-air system was simulated as a baseline to compare these configurations for assessing the energy saving potential.

Mathur, Jyotirmay↗

Transfer-Learnt Energy Models for Predicting Electricity Consumption in Buildings with Limited and Sparse Field Data

Modeling energy consumption is critical for energy-efficient utilization of the electric appliances in a building, smart grid programs (like demand-response), and many other smart home applications. State-of-the-art energy modeling techniques either rely on theoretical models, or extensive instrumentation of the building envelope to gather ``big" data to train a deep neural network. While theoretical models are often limited by their estimation accuracy, it is not always feasible to gather a significant amount of field data. In this paper, we explore transfer learning-based strategies to train much more accurate model for energy estimation when using a sparse field data. We transferred knowledge, in the form of data and parameters, from the simulation framework to the field data. We evaluated the efficacy of our approach on field data collected from six commercial buildings and our results indicate that transfer learning-based models trained over one month data can perform comparative (and in some cases better) than the state-of-the-art machine learning and deep learning solutions.

Jain, Milan↗

Dynamic Binary Complexes (DBC) as Super-Adjustable Viscosity Modifiers for Hydraulic Fracturing Fluids

In the preceding project year two, we refined three DBC formulations from a selection of over 50 different chemistries. The optimization study primarily encompassed testing for (i) reversibility extent, (ii) performance in the presence of chemical additives, (iii) adhesion and friction behavior during displacement in wellbores and pipelines, (iv) corrosion protection performance, and (v) injection performance with model fracture systems at the laboratory scale. Highly promising results obtained from all these tests signify the significant potential of DBCs in enhancing hydrocarbon recovery from unconventional reservoirs. The primary activities in the third project year included publishing experimental findings across multiple articles and conducting outreach initiatives. Throughout the year, we undertook tasks such as replicating experimental results, further optimizing various formulations and their associated experimental sets, and conducting additional tests to address missing components based on reviewer feedback and suggestions. We also explored the surfactant and friction-reduction aspects of selected formulations through drag reduction tests. In addition, we constructed an improved fracturing performance setup and performed flow injection tests. The specific DBC formulations focused on during this project period were A8/B1, A12/B5, and A10/B12. We also obtained results for additional DBC formulations and a select few commercial fracturing fluids for the purpose of comparison. Within the project's scope, we aim to enhance the experimental findings with the development of various models. The first two years focused on two key aspects: (i) the creation of a high-fidelity hydraulic fracturing model for non-Newtonian fluids to gain insights into the implementation of DBC fluids in fracking environments, and (ii) the development of a multiphase flow simulator for estimating total production, fluid saturation in the reservoir, and the creation of a fracture propagation model and kinetic Monte Carlo (kMC) models for diverse applications. In the third year, we delved into the fundamental nanostructural properties of DBCs, exploring aspects such as material chemistry, pH tunability, and control of DBC formation and stability. Subsequently, in the extension year, we conducted a systematic investigation of various building blocks containing primary, secondary, and tertiary amine functional groups to understand their impact on rheological and viscoelastic properties. Furthermore, we explored a Dissipative Particle Dynamics (DPD) model to simulate self-assembly processes with precision, creating a high-fidelity representation of relevant nanostructures. The tasks performed this year with the significant results obtained have been discussed in Section 2.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

SU-E-T-560: Monte Carlo Simulation of the Neutron Radiation Field Around a Medical 18 MV Linac

Purpose: Today the majority of radiation therapy treatments are performed at medical electron linear accelerators (linacs). The accelerated electrons are used for the generation of bremsstrahlung photons. The use of higher electron respectively photon energies has some advantages over lower energies such as the longer dose build-up. However photons with energies higher than ∼7 MeV can additionally to the interaction with bound electrons undergo inelastic reactions with nuclei. These photonuclear reactions lead to the emission of fast neutrons which contaminate the primary photon field. The neutrons might penetrate through the collimators and deliver out-of-field dose to the patient. Furthermore the materials inside the linac head as well as the air inside the treatment room get activated which might deliver dose to the medical employees even when the linac is not in operation. A detailed knowledge of these effects is essential for adequate radiation protection of the employees and an optimal patient treatment. Methods: It is a common method to study the radiation fields of such linacs by means of Monte Carlo simulations. For the investigation of the effects caused by photonuclear reactions a typical linac in high energy mode (Varian Clinac 18 MV-X) as well as the surrounding bunker were modelled and simulated using the Monte Carlo code FLUKA which includes extensive nuclear reaction and neutron transport models additional to electron-photon transport as well as capabilities for a detailed study of effective dose distributions and activation yields. Results: Neutron spectra as well as neutron effective dose distributions within the bunker were obtained, reaching up to some mSv/Gy in the patient’s plane. The results are normalized per Gy in the depth dose maximum at 10×10 cm{sup 2} field size. Therefore an absolute interpretation is possible. Conclusion: The obtained data gives a better understanding of the photonuclear reaction caused effects.

61 RADIATION PROTECTION AND DOSIMETRY↗

Reinforcement Learning Control for Buildings Co-Optimizing Energy, Comfort, and Indoor Air Quality: An Annual Assessment

Efficient control of Heating, Ventilation, and Air Conditioning (HVAC) systems is crucial for optimizing energy use and maintaining indoor comfort in buildings. Traditional control methods, such as PID control, cannot handle energy use trade-offs among multiple components in the building energy system at a supervisory level. Reinforcement learning (RL) presents a promising solution, offering adaptive and data-driven control strategies that optimize performance over time. However, RL also faces several challenges, including the conflicts encountered in co-optimizing energy savings, occupant comfort, and indoor air quality, and the requirement for extensive interactions with the environment in training. We proposed a flexible simulation platform that integrates a hybrid model for RL training and designed an RL agent to control the entire central HVAC system, focusing on co-optimizing energy consumption, thermal comfort, and indoor air quality ($\text{CO}_{2}$ and PM2.5 concentrations). Finally, we evaluated the RL agent's performance over an annual cycle. Our findings indicate that the RL agent can effectively manage the HVAC system with 14.7 % energy savings annually and balance multiple objectives, which demonstrates significant potential for improving HVAC system control and sustainability in buildings.

Guo, Fangzhou↗

An implementation of a tree code on a SIMD, parallel computer

We describe a fast tree algorithm for gravitational N-body simulation on SIMD parallel computers. The tree construction uses fast, parallel sorts. The sorted lists are recursively divided along their x, y and z coordinates. This data structure is a completely balanced tree (i.e., each particle is paired with exactly one other particle) and maintains good spatial locality. An implementation of this tree-building algorithm on a 16k processor Maspar MP-1 performs well and constitutes only a small fraction (approximately 15%) of the entire cycle of finding the accelerations. Each node in the tree is treated as a monopole. The tree search and the summation of accelerations also perform well. During the tree search, node data that is needed from another processor is simply fetched. Roughly 55% of the tree search time is spent in communications between processors. We apply the code to two problems of astrophysical interest. The first is a simulation of the close passage of two gravitationally, interacting, disk galaxies using 65,636 particles. We also simulate the formation of structure in an expanding, model universe using 1,048,576 particles. Our code attains speeds comparable to one head of a Cray Y-MP, so single instruction, multiple data (SIMD) type computers can be used for these simulations. The cost/performance ratio for SIMD machines like the Maspar MP-1 make them an extremely attractive alternative to either vector processors or large multiple instruction, multiple data (MIMD) type parallel computers. With further optimizations (e.g., more careful load balancing), speeds in excess of today's vector processing computers should be possible.

Olson, Kevin M.↗

Methodology for Constructing Reduced-Order Power Block Performance Models for CSP Applications: Preprint

The inherent variability of the solar resource presents a unique challenge for CSP systems. Incident solar irradiation can fluctuate widely over a short time scale, but plant performance must be assessed for long time periods. As a result, annual simulations with hourly (or sub-hourly) timesteps are the norm in CSP analysis. A highly detailed power cycle model provides accuracy but tends tosuffer from prohibitively long run-times; alternatively, simplified empirical models can run quickly but don?t always provide enough information, accuracy, or flexibility for the modeler. The ideal model for feasibility-level analysis incorporates both the detail and accuracy of a first-principle model with the low computational load of a regression model. The work presented in this paperproposes a methodology for organizing and extracting information from the performance output of a detailed model, then using it to develop a flexible reduced-order regression model in a systematic and structured way. A similar but less generalized approach for characterizing power cycle performance and a reduced-order modeling methodology for CFD analysis of heat transfer from electronic deviceshave been presented. This paper builds on these publications and the non-dimensional approach originally described.

14 SOLAR ENERGY↗

Analyzing the Impact of Future Weather Data on Energy Consumption in Weatherization Assistant

This study supports the mission of the U.S. Department of Energy’s Weatherization Assistance Program (WAP), which aims to increase the energy efficiency of dwellings and reduce their total residential expenditures. Specifically, we examine how projected future climate conditions may affect residential building energy performance by integrating future weather data into the National Energy Audit Tool (NEAT). Since WAP evaluates the cost-effectiveness of retrofit measures over lifespans of up to 30 years, accounting for evolving climate conditions is increasingly important. To reflect future household energy demands, this study replaces historically based Typical Meteorological Year (TMY3) weather inputs with Future Typical Meteorological Year (fTMY) datasets derived from global climate model (GCM) projections. A simulation-based framework was established to enable NEAT analysis under future weather conditions. This workflow involves converting EPW-format weather files into JSON inputs compatible with NEAT and generating degree-hour metrics needed for load calculations. The fTMY dataset used in this study was developed by Oak Ridge National Laboratory through downscaling of six GCMs under different emission scenarios and covers the period from 2020 to 2100. In contrast, the TMY3 dataset is based on historical weather data from 1961 to 1990. Simulations were conducted for benchmark single-family prototype buildings across ASHRAE climate zones 1–7, which cover all regions of the U.S. except the subarctic Zone 8 in northern Alaska, evaluating both heating and cooling loads under TMY3 and fTMY conditions. Four foundation types were tested, while heating systems were standardized, as NEAT does not differentiate thermal energy load by HVAC system type in its load calculations. Results show that fTMY weather input consistently yield lower heating loads and higher cooling loads across most locations, aligning with expected climate warming trends. Notably, colder regions such as zones 6A, 6B, and 7 experience marked reductions in heating load, while warmer and transitional zones, such as 2A (Lufkin, TX) and 3C (San Francisco, CA), have substantial increases in cooling loads. Although this study does not directly assess the performance of retrofit measures under future climate conditions, it provides a critical foundation for doing so. By quantifying shifts in baseline (i.e., pre-retrofit case) energy loads between historical and future weather files, the study highlights the importance of integrating climate-responsive data into audit tools. These findings will inform future efforts to evaluate the long-term effectiveness and cost-effectiveness of weatherization measures under changing climate conditions.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Center for Center for Technology for Advanced Scientific Component Software (TASCS)

A resounding success of the Scientific Discovery through Advanced Computing (SciDAC) program is that high-performance computational science is now universally recognized as a critical aspect of scientific discovery [71], complementing both theoretical and experimental research. As scientific communities prepare to exploit unprecedented computing capabilities of emerging leadership-class machines for multi-model simulations at the extreme scale [72], it is more important than ever to address the technical and social challenges of geographically distributed teams that combine expertise in domain science, applied mathematics, and computer science to build robust and flexible codes that can incorporate changes over time. The Center for Technology for Advanced Scientific Component Software (TASCS)1 tackles these these issues by exploiting component-based software development to facilitate collaborative high-performance scientific computing.

97 MATHEMATICS AND COMPUTING↗

Assessment of Literature Related to Combustion Appliance Venting Systems

In many residential building retrofit programs, air tightening to increase energy efficiency is constrained by concerns about related impacts on the safety of naturally vented combustion appliances. Tighter housing units more readily depressurize when exhaust equipment is operated, making combustion appliances more prone to backdraft or spillage. Several test methods purportedly assess the potential for depressurization-induced backdrafting and spillage, but these tests are not robustly reliable and repeatable predictors of venting performance, in part because they do not fully capture weather effects on venting performance. The purpose of this literature review is to investigate combustion safety diagnostics in existing codes, standards, and guidelines related to combustion appliances. This review summarizes existing combustion safety test methods, evaluations of these test methods, and also discusses research related to wind effects and the simulation of vent system performance. Current codes and standards related to combustion appliance installation provide little information on assessing backdrafting or spillage potential. A substantial amount of research has been conducted to assess combustion appliance backdrafting and spillage test methods, but primarily focuses on comparing short-term (stress) induced tests and monitoring results. Monitoring, typically performed over one week, indicated that combinations of environmental and house operation characteristics most conducive to combustion spillage were rare. Research, to an extent, has assessed existing combustion safety diagnostics for house depressurization, but the objectives of the diagnostics, both stress and monitoring, are not clearly defined. More research is also needed to quantify the frequency of test “failure” occurrence throughout the building stock and assess the statistical effects of weather (especially wind) on house depressurization and in turn on combustion appliance venting. Incorporating weather variations and house ventilation system characteristics in existing simulation software may assist such analyses and with developing a more reliable diagnostic for use on-site.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗

Energy Impacts of Nonlinear Behavior of PCM When Applied into Building Envelope

Presented at the ASME 2012 6th International Conference on Energy Sustainability & 10th Fuel Cell Science, Engineering and Technology Conference on July 23-26, 2012, this study analyzes the effects a nonlinear enthalpy profile has on thermal performance and expected energy benefits for PCM-enhanced insulation.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Active Magnetic Regenerative Liquefier

This final report for the DOE Project entitled Active Magnetic Regenerative Liquefier (AMRL) funded under Grant DE-FG36-08GO18064 to Heracles Energy Corporation d.b.a. Prometheus Energy (Heracles/Prometheus) describes an active magnetic regenerative refrigerator (AMRR) prototype designed and built during the period from July 2008 through May 2011. The primary goal of this project was to make significant technical advances toward highly efficient liquefaction of hydrogen. Conventional hydrogen liquefiers at any scale have a maximum FOM of ~0.35 due primarily to the intrinsic difficulty of rapid, efficient compression of either hydrogen or helium working gases. Numerical simulation modeling of high performance AMRL designs indicates certain designs have promise to increase thermodynamic efficiency from a FOM of ~0.35 toward ~0.5 to ~0.6. The technical approach was the use of solid magnetic working refrigerants cycled in and out of high magnetic fields to build an efficient active regenerative magnetic refrigeration module providing cooling power for AMRL. A single-stage reciprocating AMRR with a design temperature span from ~290 K to ~120 K was built and tested with dual magnetic regenerators moving in and out of the conductively-cooled superconducting magnet subsystem. The heat transfer fluid (helium) was coupled to the process stream (refrigeration/liquefaction load) via high performance heat exchangers. In order to maximize AMRR efficiency a helium bypass loop with adjustable flow was incorporated in the design because the thermal mass of magnetic refrigerants is higher in low magnetic field than in high magnetic field. Heracles/Prometheus designed experiments to measure AMRR performance under a variety of different operational parameters such as cycle frequency, magnetic field strength, heat transfer fluid flow rate, amount of bypass flow of the heat transfer fluid while measuring work input, temperature span, cooling capability as a function of cold temperature as a function of the amount of bypass flow of the heat transfer fluid. The operational AMRR prototype can be used to answer key questions such as the best recipe for multiple layers of different magnetic refrigerants in one or more integrated regenerators with varying amounts of bypass flow of the heat transfer fluid. Layered regenerators are necessary to span the AMRR range from 290 K to 120K. Our AMRR performance simulation model predicts that ~10-15 % of bypass flow should significantly improve the thermodynamic performance. Initial results obtained with regenerators made of gadolinium spheres were very encouraging; a temperature span of ~ 50 K (between 295K and 245 K) across both regenerators was achieved with zero bypass flow of the heat transfer fluid and with the magnetic field strength of ~4 T.

42 ENGINEERING↗

Developing Urban Air Mobility Vehicle Models to Support Air Traffic Management Concept Development

To support Urban Air Mobility (UAM) research efforts at NASA, the Airspace Target Generator (ATG) software used in the FutureFlight Central (FFC) air traffic control tower simulator is undergoing updates to support physics-based UAM vehicle models. A process was developed to integrate UAM vertical takeoff and landing (VTOL) aircraft into the fixed-wing ATG modeling environment without significant change to the underlying equations of motion and vehicle model database. The VTOL aircraft models were converted from a six degrees-of-freedom (6-DOF) representation into a four degrees-of-freedom (4-DOF) representation for integration within ATG. Three vehicle designs from the NASA Revolutionary Vertical-Lift Technologies (RVLT) project were selected: a lift-plus-cruise (LPC) aircraft model and quadrotor, electric-powered (QEP) 1-seater and 6-seater models. With the LPC model comprised of a nonlinear force and moment build-up, and the QEP models comprised of linearized stability derivatives, two separate processes were developed to convert the lift, drag, and propulsion characteristics of each model into the ATG model database. Key aircraft performance characteristics including climb, cruise, and descent performance were preserved during the conversion process. Because ATG simulates fixed-wing aircraft through ground taxi and takeoff to approach and landing, acceleration command algorithms were developed to model the vertical takeoff and vertical landing phase of UAM operations. A strategy was then developed to transition the aircraft model to- and from- the new control mode.

urban air mobility↗

Deep reinforcement learning with online data augmentation to improve sample efficiency for intelligent HVAC control

Deep Reinforcement Learning (DRL) has started showing success in real-world applications such as building energy optimization. Much of the research in this space utilized simulated environments to train RL-agent in an offline mode. Very few research have used DRL-based control in real-world systems due to two main reasons: 1) sample efficiency challenge---DRL approaches need to perform a lot of interactions with the environment to collect sufficient experiences to learn from, which is difficult in real systems, and 2) comfort or safety related constraints---user's comfort must never or at least rarely be violated. In this work, we propose a novel deep Reinforcement Learning framework with online Data Augmentation (RLDA) to address the sample efficiency challenge of real-world RL. We used a time series Generative Adversarial Network (TimeGAN) architecture as a data generator. We further evaluated the proposed RLDA framework using a case study of an intelligent HVAC control. With a ≈28% improvement in the sample efficiency, RLDA framework lays the way towards increased adoption of DRL-based intelligent control in real-world building energy management systems.

Kurte, Kuldeep↗