Search NASA⌕ Search

SEARCH · Search NASA

Results for “Real Time Optimization”

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 1,567 records · Page 87

Integrated simulations for fusion research in the 2030's time frame (white paper outline)

This white paper presents the rationale for developing a community-wide capability for whole-device modeling, and advocates for an effort with the expectation of persistence: a long-term programmatic commitment, and support for community efforts. Statement of 2030 goal (two suggestions): (a) Robust integrated simulation tools to aid real-time experimental discharges and reactor designs by employing a hierarchy in fidelity of physics models. (b) To produce by the early 2030s a capability for validated, predictive simulation via integration of a suite of physics models from moderate through high fidelity, to understand and plan full plasma discharges, aid in data interpretation, carry out discovery science, and optimize future machine designs. We can achieve this goal via a focused effort to extend current scientific capabilities and rigorously integrate simulations of disparate physics into a comprehensive set of workflows.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

A data acquisition system for longitudinal beam properties in a rapid cycling synchrotron

A longitudinal beam properties, data acquisition system has been commissioned to operate in the Fermilab booster ring. This system captures real time information including beam synchronous phase, bunch length, and coupled bunch instability amplitudes as the beam is accelerated from 400 MeV to 8 GeV in 33 ms. The system uses an off-the-shelf Tektronix oscilloscope running Labview software and a synchronous pulse generator. This paper describes the hardware configuration and the software configuration used to optimize the data processing rate. The Fermilab Booster is currently pushed to its intensity limit due to the high demand from different experiments for protons. This will continue into the era of the intensity frontier. While intensity is the most critical parameter, emittance must be controlled as well. Longitudinal emittance is an important parameter for optimizing the slip stacking process in the Main Injector. Longitudinal instabilities can spoil the emittance, so it is important to monitor bunch lengths and bunch phase oscillations in the Booster. The Booster presents a special challenge when it comes to measuring beam properties. The acceleration ramp is short (33ms), and the RF frequency varies from 38-52.8 MHz. Standard instrumentation like spectrum analyzers and network analyzers cannot effectively measure beam properties, because they cannot track the changing frequencies. Early instrumentation involved a series of tracking oscillators, mixers, and diode detectors that would require vigilant calibration. There was a demand for a more robust means of gathering beam parameters.

43 PARTICLE ACCELERATORS↗

In-situ X-Ray Analysis of Rapid Thermal Processing for Thin-Film Solar Cells: Closing the Gap between Production and Laboratory Efficiency

For materials synthesis, it is well known that the material final state may not reach equilibrium and depends on the synthetic process. In particular, processes that quickly remove the available energy from the material may leave it in a metastable state and the metastability may actually impart desirable functional properties. By its very nature, Rapid thermal processing (RTP) is ideally suited to produce such metastable materials. However, metastability and the dynamics of reaching a metastable state are poorly understood, since this is best accomplished through in situ monitoring. In this regard, RTP is particularly challenging as the processing time are very short (seconds to minutes). As a result, there is only poor understanding, and hence use, of RTP in industry. This is potentially a cost-increasing limitation, because RTP can decrease cost by decreasing processing time, and as such, increase throughput and decrease the total thermal budget of processing - a significant cost. RTP is already being used for key processing steps in PV technologies. With silicon wafer PV, it is used for establishing electrical contact between the Ag metal grid and the silicon (known as firing). In this process, a silicon wafer with deposited metal/frit in a grid pattern is heated rapidly to temperatures between 750 and 800 ºC. The processing time when the temperature is held above 600ºC is short (<5 seconds). This process has historically been optimized empirically and it is unclear how the thermal processing affects formation of the final contact between the metal and the silicon. In the case of thin-film PV, RTP has been demonstrated in the process of making absorber layers, i.e. CIGS and CZTS. Use of RTP can reduce the processing time from 10s of minutes to seconds, reducing the thermal budget and increasing the throughput significantly. The conversion from precursor material to final PV material is not well understood, and most of the process optimization is done empirically. Again, the proposed in situ monitoring will provide insight into the conversion process. We have developed a unique instrument for in situ monitoring of the reaction process during RTP. The most fundamental property to monitor is the atomic arrangement (for solids, crystalline structure) and thus we have focused on this property. The chamber was designed to collect X-ray diffraction (XRD) for crystal phase determination, although it has lent insight in the formation of liquids. We designed and constructed both inert (air) and reactive (S/Se) versions of this chamber. We have applied the RTP-XRD methodology to: silicon solar cell electrical contact formation; the reaction sequence for forming CIGS PV thin films from copper selenide and indium gallium selenide precursors; the phase sequence for forming HC(NH2)2PbI3 (FAPbI3) and CH3NH3PbI3 (MAPbI3) hybrid-perovskite PV thin films. The RTP instrument enables real-time collection of X-ray diffraction data with intervals as short as 100 ms, and while heating at ramp rates as high 100 ºCs-1 up to 1200 ºC. The system is portable and can be installed on a synchrotron beamline. The RTP reactor consists of a metallic body (front cap assembly, central block and end cap assembly) that houses a quartz reaction chamber. The sample is secured inside the reaction chamber on a sample holder connected to the front-cap assembly. Modifications of the chamber were made for safely handling sulfur and selenium vapor environments. A Se/S source holder was added to generate a Se/S over pressure in the chamber on heating and the entrance window is made wider in order to accommodate the detachable additional holder. Se/S are toxic in nature and therefore a proper Se/S trapping is used before releasing the gases to exhaust. Screen-printing provides an economically attractive means for making Ag electrical contacts to Si solar cells, but the use of Ag substantiates a significant manufacturing cost, and the glass frit used in the paste to enable contact formation contains Pb. To achieve optimal electrical performance and to develop pastes with alternative, abundant, and non-toxic materials requires understanding the contact formation process during firing. We use the RTP-XRD chamber during firing of Ag-Si contacts to reveal the reaction sequence for contact formation. The findings show that between 500 ºC and 650 ºC PbO in the frit etches the SiNx antireflective-coating on the solar cell, exposing the Si surface. Then, above 650 ºC, Ag+ dissolves into the molten glass frit – key for enabling deposition of metallic Ag on the emitter surface and precipitation of Ag nanocrystals within the glass. This clarifies contact formation mechanisms and can be used to suggest approaches for development of inexpensive, nontoxic solar cell contacting pastes. We used the RTP/XRD chamber to study the formation of CuInGaS/Se (CIGS) PV films by in-situ selenization of pre-deposited metal stacks and annealing of pre-deposited Cu-In-Ga-Se precursors. During the metal stack selenization, we observed that Se first reacts with the metals and forms selenides. We observed formation of intermetallics such as Cu2In and binary selenides during heating. On further annealing in a Se atmosphere, CuInSe2 and CuGaSe2 forms which on heating up to 550 °C reacts further to form the CIGS phase. On prolonged exposure to 550 °C for more than 5 mins or higher temperature, we see separation into Ga-rich and Ga-depleted CIGS phases. On introducing a holding step at 400 °C for 10 min during the annealing ramp, the CIGS phase (forms between 500-550 °C), appears to be less prone to decomposition into high-Ga and low-Ga phases during holding at peak temperature. Lead halide perovskites have emerged as successful optoelectronic materials with high photovoltaic power conversion efficiencies and low material cost. However, substantial challenges remain in the scalability, stability and fundamental understanding of the materials. We demonstrated the application of radiative thermal annealing, an easily scalable processing method, for synthesizing formamidinium lead iodide (FAPbI3) perovskite solar absorbers. Devices fabricated from films formed via radiative thermal annealing have equivalent efficiencies to those annealed using a conventional hotplate. By coupling results from in situ x-ray diffraction using a radiative thermal annealing system with device performances, we mapped the processing phase space of FAPbI3 and corresponding device efficiencies. Our map of processing-structure-performance space suggests the commonly used FAPbI3 annealing time, 10 min at 170 °C, can be significantly reduced to 40 s at 170 °C without affecting the photovoltaic performance. Applying the Johnson-Mehl-Avrami model to the decomposition into PbI2, the activation energy was determined. We also presented a well-controlled, manufacturing relevant annealing method, radiative thermal annealing, for the methylammonium lead triiodide films formed by solvent engineering process, with dimethylformamide (DMF) and dimethyl sulfoxide as solvent and diethyl ether as the antisolvent. Radiative thermal annealing can produce high quality perovskite films, evidenced by high efficiency solar cell devices, in a shorter time than the widely-used hotplate annealing. Using in-situ x-ray diffraction during the radiative annealing, we show that the role of the antisolvent is not to form an intermediate compound (a methylammonium iodide, lead iodide, and dimethyl sulfoxide compound) by washing of the main solvent (DMF) but to achieve a pin-hole free, uniform film of methylammonium lead iodide. We find that having the intermediate compound does not guarantee a high quality (pin-hole free) perovskite film. Our study was extended to reveal the effect of annealing environment, annealing temperature profile, and deposition substrate.

36 MATERIALS SCIENCE↗

Cation stoichiometry optimization of SrTiO{sub 3} (110) thin films with atomic precision in homogeneous molecular beam epitaxy

We study the evolution of surface reconstructions on SrTiO{sub 3} (110) determined by cation concentration ratio and find it detectable by high energy electron diffraction (RHEED) even at high temperature up to 800 deg. C. The evaporation rate of Sr and Ti sources can be calibrated precisely by monitoring RHEED patterns in situ and in real time during the extended homogeneous oxide molecule beam epitaxy growth of SrTiO{sub 3} (110) film. High quality film is obtained with deviation of cation stoichiometry less than 0.5%.

36 MATERIALS SCIENCE↗

Interrogation of bimetallic particle oxidation in three dimensions at the nanoscale

An understanding of bimetallic alloy oxidation is key to the design of hollow-structured binary oxides and the optimization of their catalytic performance. However, one roadblock encountered in studying these binary oxide systems is the difficulty in describing the heterogeneities that occur in both structure and chemistry as a function of reaction coordinate. This is due to the complexity of the three-dimensional mosaic patterns that occur in these heterogeneous binary systems. By combining real-time imaging and chemical-sensitive electron tomography, we show that it is possible to characterize these systems with simultaneous nanoscale and chemical detail. We find that there is oxidation-induced chemical segregation occurring on both external and internal surfaces. Additionally, there is another layer of complexity that occurs during the oxidation, namely that the morphology of the initial oxide surface can change the oxidation modality. As a result, this work characterizes the pathways that can control the morphology in binary oxide materials.

36 MATERIALS SCIENCE↗

Machine Learning-Enabled Image Classification for Automated Electron Microscopy

Abstract Traditionally, materials discovery has been driven more by evidence and intuition than by systematic design. However, the advent of “big data” and an exponential increase in computational power have reshaped the landscape. Today, we use simulations, artificial intelligence (AI), and machine learning (ML) to predict materials characteristics, which dramatically accelerates the discovery of novel materials. For instance, combinatorial megalibraries, where millions of distinct nanoparticles are created on a single chip, have spurred the need for automated characterization tools. This paper presents an ML model specifically developed to perform real-time binary classification of grayscale high-angle annular dark-field images of nanoparticles sourced from these megalibraries. Given the high costs associated with downstream processing errors, a primary requirement for our model was to minimize false positives while maintaining efficacy on unseen images. We elaborate on the computational challenges and our solutions, including managing memory constraints, optimizing training time, and utilizing Neural Architecture Search tools. The final model outperformed our expectations, achieving over 95% precision and a weighted F-score of more than 90% on our test data set. This paper discusses the development, challenges, and successful outcomes of this significant advancement in the application of AI and ML to materials discovery.

Materials Science↗

Large Scale GPU Accelerated PPMLR-MHD Simulations for Space Weather Forecast

PPMLR-MHD is a new magnetohydrodynamics (MHD) model used to simulate the interactions of the solar wind with the magnetosphere, which has been proved to be the key element of the space weather cause-and-effect chain process from the Sun to Earth. Compared to existing MHD methods, PPMLR-MHD achieves the advantage of high order spatial accuracy and low numerical dissipation. However, the accuracy comes at a cost. On one hand, this method requires more intensive computation. On the other hand, more boundary data is subject to be transferred during the process of simulation. In this work, we present a parallel hybrid solution of the PPMLR-MHD model implemented using the computing capabilities of both CPUs and GPUs. We demonstrate that our optimized implementation alleviates the data transfer overhead by using GPU Direct technology and can scale up to 151 processes and achieve significant performance gains by distributing the workload among the CPUs and GPUs on Titan at Oak Ridge National Laboratory. The performance results show that our implementation is fast enough to carry out highly accurate MHD simulations in real time.

Computer Science↗

Profiling optimization for big data transfer over dedicated channels

The transfer of big data is increasingly supported by dedicated channels in high-performance networks, where transport protocols play an important role in maximizing applicationlevel throughput and link utilization. The performance of transport protocols largely depend on their control parameter settings, but it is prohibitively time consuming to conduct an exhaustive search in a large parameter space to find the best set of parameter values. We propose FastProf, a stochastic approximation-based transport profiler, to quickly determine the optimal operational zone of a given data transfer protocol/method over dedicated channels. We implement and test the proposed method using both emulations based on real-life performance measurements and experiments over physical connections with short (2 ms) and long (380 ms) delays. Both the emulation and experimental results show that FastProf significantly reduces the profiling overhead while achieving a comparable level of end-to-end throughput performance with the exhaustive search-based approach.

Yun, D.↗

Dynamic System Scaling Application to Accelerated Nuclear Fuel Testing

The development of nuclear fuel and materials requires a continuous effort to investigate the acute and prolonged effects of irradiation, thermal-material stress, chemical change, or other conceivable damage mechanics acquired during normal operation or accident scenarios throughout its lifetime. As in-core fuel property measurement techniques advance to support in the real-time, non-invasive, and enhanced accuracy realm, it is the future of fuel development to pursue to higher degrees of control, predictability of integral test behavior via separate effects test (SET), and shorter test time intervals. The fuel development life cycle from initial concept to commercial licensing is approximated to be 20 years and current literature suggests by optimizing fuel performance codes with SETs, the process could possibly be compressed to 5 to 10 years. Recently, a research group in the Idaho National Laboratory (INL) is testing reduced scale fuel rods and increased power density to accelerate evolution of fuel phenomena in metallic fuels. In support of nuclear fuel rod development, compressing fuel test process, and accelerating fuel phenomena, it was the purpose of this study to investigate nuclear fuel performance phenomena via literature review and effectively scale the initial conditions, boundary conditions, and geometric properties to describe the time-dependent response including to fuel burnup, thermal and mechanical stress, transmutations and inter-diffusion, and other relevant observed phenomena. The study was based on BISON simulations of historic EBR-II metallic fuel experiments and Dynamical System Scaling (DSS) method are utilized to assess effects of scaling activity including fuel phenomena acceleration and calculations of time-dependent distortions. The research successfully scaled metallic fuel phenomena, accelerated fuel testing, and assessed the distortions for each scaled case derived.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

Beyond PID Controllers: PPO with Neuralized PID Policy for Proton Beam Intensity Control in Mu2e

We introduce a novel Proximal Policy Optimization (PPO) algorithm aimed at addressing the challenge of maintaining a uniform proton beam intensity delivery in the Muon to Electron Conversion Experiment (Mu2e) at Fermi National Accelerator Laboratory (Fermilab). Our primary objective is to regulate the spill process to ensure a consistent intensity profile, with the ultimate goal of creating an automated controller capable of providing real-time feedback and calibration of the Spill Regulation System (SRS) parameters on a millisecond timescale. We treat the Mu2e accelerator system as a Markov Decision Process suitable for Reinforcement Learning (RL), utilizing PPO to reduce bias and enhance training stability. A key innovation in our approach is the integration of a neuralized Proportional-Integral-Derivative (PID) controller into the policy function, resulting in a significant improvement in the Spill Duty Factor (SDF) by 13.6%, surpassing the performance of the current PID controller baseline by an additional 1.6%. This paper presents the preliminary offline results based on a differentiable simulator of the Mu2e accelerator. It paves the groundwork for real-time implementations and applications, representing a crucial step towards automated proton beam intensity control for the Mu2e experiment.

43 PARTICLE ACCELERATORS↗

Hi-Q Rotor - Low Wind Speed Technology

The project objective was to optimize the performance of the Hi-Q Rotor. Early research funded by the California Energy Commission indicated the design might be advantageous over state-of-the-art turbines for collecting wind energy in low wind conditions. The Hi-Q Rotor is a new kind of rotor targeted for harvesting wind in Class 2, 3, and 4 sites, and has application in areas that are closer to cities, or 'load centers.' An advantage of the Hi-Q Rotor is that the rotor has non-conventional blade tips, producing less turbulence, and is quieter than standard wind turbine blades which is critical to the low-wind populated urban sites. Unlike state-of-the-art propeller type blades, the Hi-Q Rotor has six blades connected by end caps. In this phase of the research funded by DOE's Inventions and Innovation Program, the goal was to improve the current design by building a series of theoretical and numeric models, and composite prototypes to determine a best of class device. Development of the rotor was performed by aeronautical engineering and design firm, DARcorporation. From this investigation, an optimized design was determined and an 8-foot diameter, full-scale rotor was built and mounted using a Bergey LX-1 generator and furling system which were adapted to support the rotor. The Hi-Q Rotor was then tested side-by-side against the state-of-the-art Bergey XL-1 at the Alternative Energy Institute's Wind Test Center at West Texas State University for six weeks, and real time measurements of power generated were collected and compared. Early wind tunnel testing showed that the cut-in-speed of the Hi-Q rotor is much lower than a conventional tested HAWT enabling the Hi-Q Wind Turbine to begin collecting energy before a conventional HAWT has started spinning. Also, torque at low wind speeds for the Hi-Q Wind Turbine is higher than the tested conventional HAWT and enabled the wind turbine to generate power at lower wind speeds. Based on the data collected, the results of our first full-scale prototype wind turbine proved that higher energy can be captured at lower wind speeds with the new Hi-Q Rotor. The Hi-Q Rotor is almost 15% more productive than the Bergey from 6 m/s to 8 m/s, making it ideal in Class 3, 4, and 5 wind sites and has application in the critical and heretofore untapped areas that are closer to cities, 'load centers,' and may even be used directly in urban areas. The additional advantage of the Hi-Q Rotor's non-conventional blade tips, which eliminates most air turbulence, is noise reduction which makes it doubly ideal for populated urban areas. Hi-Q Products recommends one final stage of development to take the Hi-Q Rotor through Technology Readiness Levels 8-9. During this stage of development, the rotor will be redesigned to further increase efficiency, match the rotor to a more suitable generator, and lower the cost of manufacturing by redesigning the structure to allow for production in larger quantities at lower cost. Before taking the rotor to market and commercialization, it is necessary to further optimize the performance by finding a better generator and autofurling system, ones more suitable for lower wind speeds and rpms should be used in all future testing. The potential impact of this fully developed technology will be the expansion and proliferation of energy renewal into the heretofore untapped Class 2, 3, 4, and 5 Wind Sites, or the large underutilized sites where the wind speed is broken by physical features such as mountains, buildings, and trees. Market estimates by 2011, if low wind speed technology can be developed are well above: 13 million homes, 675,000 commercial buildings, 250,000 public facilities. Estimated commercial exploitation of the Hi-Q Rotor show potential increase in U.S. energy gained through the clean, renewable wind energy found in low and very low wind speed sites. This new energy source would greatly impact greenhouse emissions as well as the public sector's growing energy demands.

17 WIND ENERGY↗

Benchtop Autonomous Electrochemical Characterization System for Combinatorial Thin-Film Solid Oxide Electrodes

The design of materials for electrochemical energy conversion is complicated by a vast search space of candidate materials and multifaceted property requirements: multicarrier conductivity, stability, and catalytic activity are all necessary but rarely intersect. Although self-driving laboratories are rapidly rising to address such material optimization problems, the required infrastructure for integrated, large-scale robotic facilities can be cost-prohibitive. Here we develop and evaluate a closed-loop measurement system for efficient screening of proton-conducting oxide electrodes for ceramic fuel cells and electrolyzers, building on top of an existing benchtop instrument and integrating techniques for rapid impedance measurement and automated analysis. This system exemplifies a “minimum viable” self-driving implementation that can deliver substantial benefits with relatively simple infrastructure. Combinatorial thin-film microelectrode libraries are characterized with a recently developed joint time-domain and frequency-domain impedance measurement technique, which provides an order-of-magnitude acceleration relative to conventional impedance spectroscopy. The distribution of relaxation times is extracted from impedance data and analyzed without human intervention. These results feed an active learning and Bayesian optimization process that learns to predict electrochemical impedance as a function of material composition, measurement temperature, oxygen partial pressure, and electrical bias, which further reduces the screening time by tenfold with optimized experimental sequences. We apply this system to Ba⁡(Co,Fe,Zr,Y)⁢O 3−𝛿 combinatorial libraries and evaluate its effectiveness for learning material property trends and optimizing expensive-to-evaluate properties such as activation energy. This offers insights into key methodological aspects of practical autonomous experimentation, including surrogate model validation, cost-aware acquisition functions, and high-throughput data interpretation. Our results demonstrate the efficacy of the system for rapidly gathering information, but also highlight real-world experimental challenges of thin-film degradation and numerical instability in surrogate models.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Rotating Dual-Wire Beam Profile Monitor Optimized for Use in Merged-Beams Experiments

A rotating dual-wire beam proile monitor based upon a modified National Electrostatics Corporation Model BPM80 beam profile monitor is described. The device can measure beam profiles in two perpendicular directions (horizontal and vertical) in each of two pseudoplanes that are situated along the beam axis and are separated by a distance of 6.0 cm. The output signal from the device is analyzed in real time to yield horizontal and vertical beam profiles and to calculate the divergence of a particle beam that traverses the device. This set-up is well-suited for merged-beams experiments where one beam is tuned to saved profiles from a second beam in order to minimize the merge angle and beam divergences while maximizing the beam-beam overlaps.

Seely, D. G.↗

Machine Learning Classical Interatomic Potentials for Molecular Dynamics from First-Principles Training Data

The ever-increasing power of modern super-computers, along with the availability of highly scalable atomistic simulation codes, has begun to revolutionize predictive modeling of materials. In particular, molecular dynamics (MD) has led to breakthrough advances in diverse fields, including tribology, catalysis, sensing, and nanoparticle self-assembly. Furthermore, recent integration of MD simulations with X-ray characterization has demonstrated promise in real-time 3-D characterization of materials on the atomic scale. The popularity of MD is driven by its applicability at disparate length/time scales, ranging from ab initio MD (hundreds of atoms and tens of picoseconds) to all-atom classical MD (millions of atoms over microseconds), and coarse-grained (CG) models (micrometers and tens of microseconds). Nevertheless, a substantial gap persists between AIMD, which is highly accurate but restricted to extremely small sizes, and those based on classical force fields (atomistic and CG) with limited accuracy but access to larger length/time scales. The accuracy and predictive power of classical MD simulations is dictated by the empirical force fields, and their capability to capture the relevant physics. Here, we discuss some of our recent work on the use of machine learning (ML) to combine the accuracy and flexibility of electronic structure calculations with the speed of classical potentials. Our ML framework attempts to bridge the significant gulf that exists between the handful of research groups that develop new interatomic potential models (often requiring several years of effort), and the increasingly large user community from academia and industry that applies these models. Our data-driven approach represents significant departure from the status quo and involves several steps including generation and manipulation of extensive training data sets through electronic structure calculations, defining novel potential functional forms, employing state-of-the-art ML algorithms to formulate highly optimized training procedures, and subsequently developing user-friendly workflow tools integrating these algorithms on high-performance computers (HPCs). Our ML approach shows marked success in developing force fields for a wide range of materials from metals, oxides, nitrides, and heterointerfaces to two-dimensional (2D) materials.

Chan, Henry↗

Midwest Consortium for Wind Turbine Reliability and Optimization

This report provides an overview of the efforts aimed to establish a student focused laboratory apparatus that will enhance Purdue's ability to recruit and train students in topics related to the dynamics, operations and economics of wind turbines. The project also aims to facilitate outreach to students at Purdue and in grades K-12 in the State of Indiana by sharing wind turbine operational data. For this project, a portable wind turbine test apparatus was developed and fabricated utilizing an AirX 400W wind energy converter. This turbine and test apparatus was outfitted with an array of sensors used to monitor wind speed, turbine rotor speed, power output and the tower structural dynamics. A major portion of this project included the development of a data logging program used to display real-time sensor data and the recording and creation of output files for data post-processing. The apparatus was tested in an open field to subject the turbine to typical operating conditions and the data acquisition system was adjusted to obtain desired functionality to facilitate use for student projects in existing courses offered at Purdue University and Indiana University. Data collected using the data logging program is analyzed and presented to demonstrate the usefulness of the test apparatus related to wind turbine dynamics and operations.

17 WIND ENERGY↗

Learning in continuous action space for developing high dimensional potential energy models

Reinforcement learning (RL) approaches that combine a tree search with deep learning have found remarkable success in searching exorbitantly large, albeit discrete action spaces, as in chess, Shogi and Go. Many real-world materials discovery and design applications, however, involve multi-dimensional search problems and learning domains that have continuous action spaces. Exploring high-dimensional potential energy models of materials is an example. Traditionally, these searches are time consuming (often several years for a single bulk system) and driven by human intuition and/or expertise and more recently by global/local optimization searches that have issues with convergence and/or do not scale well with the search dimensionality. Here, in a departure from discrete action and other gradient-based approaches, we introduce a RL strategy based on decision trees that incorporates modified rewards for improved exploration, efficient sampling during playouts and a “window scaling scheme" for enhanced exploitation, to enable efficient and scalable search for continuous action space problems. Using high-dimensional artificial landscapes and control RL problems, we successfully benchmark our approach against popular global optimization schemes and state of the art policy gradient methods, respectively. We demonstrate its efficacy to parameterize potential models (physics based and high-dimensional neural networks) for 54 different elemental systems across the periodic table as well as alloys. We analyze error trends across different elements in the latent space and trace their origin to elemental structural diversity and the smoothness of the element energy surface. Broadly, our RL strategy will be applicable to many other physical science problems involving search over continuous action spaces.

36 MATERIALS SCIENCE↗

A Novel Event Detection Method Using PMU Data with High Precision

To take full advantage of the considerably high reporting rate of phasor measurement units (PMU) data, this paper develops a novel PMU-based event detection methodology. Considering the huge amount of streaming PMU data, a data compression algorithm, swinging door trending (SDT), is first used to compress the PMU data and generate multiple compression intervals. Then dynamic programming is utilized to solve the optimization problem, which is recursively constituted by a score function. Based on predefined PMU event rules, dynamic programming merges adjacent compression intervals with the same slope direction. Finally, all the PMU event features are characterized. A conventional wavelet-based event detection method is compared with the developed dynamic programming based SDT (DPSDT) method. Numerical simulations on the real-time and synthetic PMU data show that the DPSDT method can accurately detect the start-time of an event and the event placement with relatively high precision. Furthermore, the PMU event features, including the magnitude and duration of strokes, are characterized.

24 POWER TRANSMISSION AND DISTRIBUTION↗

A Theoretical Open Architecture Framework and Technology Stack for Digital Twins in Energy Sector Applications

Digital twin is often viewed as a technology that can assist engineers and researchers make data-driven system and network-level decisions. Across the scientific literature, digital twins have been consistently theorized as a strong solution to facilitate proactive discovery of system failures, system and network efficiency improvement, system and network operation optimization, among others. With their strong affinity to the industrial metaverse concept, digital twins have the potential to offer high-value propositions that are unique to the energy sector stakeholders to realize the true potential of physical and digital convergence and pertinent sustainability goals. Although the technology has been known for a long time in theory, its practical real-world applications have been so far limited, nevertheless with tremendous growth projections. In the energy sector, there have been theoretical and lab-level experimental analysis of digital twins but few of those experiments resulted in real-world deployments. There may be many contributing factors to any friction associated with real-world scalable deployment in the energy sector such as cost, regulatory, and compliance requirements, and measurable and comparable methods to evaluate performance and return on investment. Those factors can be potentially addressed if the digital twin applications are built on the foundations of a scalable and interoperable framework that can drive a digital twin application across the project lifecycle: from ideation to theoretical deep dive to proof of concept to large-scale experiment to real-world deployment at scale. This paper is an attempt to define a digital twin open architecture framework that comprises a digital twin technology stack (D-Arc) coupled with information flow, sequence, and object diagrams. Those artifacts can be used by energy sector engineers and researchers to use any digital twin platform to drive research and engineering. This paper also provides critical details related to cybersecurity aspects, data management processes, and relevant energy sector use cases.

Gourisetti, Sri, Nikhil Gupta (ORCID:0000000188778↗