SEARCH · Engineering Papers
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.
In this paper, upper bounds to the average maximum-likelihood bit error probability of serially concatenated coding schemes are derived, design guidelines for the outer and inner codes that maximize the interleaver gain are presented, and finally, a highly-performing iterative decoding algorithm is proposed.
This paper derives upper bounds on the expected number of search tree nodes visited during an m-solution backtracking search, a search which terminates after some preselected number m problem solutions are found. The search behavior is assumed to have a general probabilistic structure. The results are stated in terms of node expansion and contraction. A visited search tree node is said to be expanding if the mean number of its children visited by the search exceeds 1 and is contracting otherwise. It is shown that if every node expands, or if every node contracts, then the number of search tree nodes visited by a search has an upper bound which is linear in the depth of the tree, in the mean number of children a node has, and in the number of solutions sought. Also derived are bounds linear in the depth of the tree in some situations where an upper portion of the tree contracts (expands), while the lower portion expands (contracts). While previous analyses of 1-solution backtracking have concluded that the expected performance is always linear in the tree depth, the model allows superlinear expected performance.
This paper shows that if one is provided with a loss function, it can be used in a natural way to specify a distance measure quantifying the similarity of any two supervised learning algorithms, even non-parametric algorithms. Intuitively, this measure gives the fraction of targets and training sets for which the expected performance of the two algorithms differs significantly. Bounds on the value of this distance are calculated for the case of binary outputs and 0-1 loss, indicating that any two learning algorithms are almost exactly identical for such scenarios. As an example, for any two algorithms A and B, even for small input spaces and training sets, for less than 2e(-50) of all targets will the difference between A's and B's generalization performance of exceed 1%. In particular, this is true if B is bagging applied to A, or boosting applied to A. These bounds can be viewed alternatively as telling us, for example, that the simple English phrase 'I expect that algorithm A will generalize from the training set with an accuracy of at least 75% on the rest of the target' conveys 20,000 bytes of information concerning the target. The paper ends by discussing some of the subtleties of extending the distance measure to give a full (non-parametric) differential geometry of the manifold of learning algorithms.
In the evolving landscape of high-performance computing, especially to fight the end of Moore’s Law and Dennard’s Scaling, the ability to shift between compute-bound and memory-bound states is critical for enhancing adaptability and flexibility to diverse system and domain-specific architectures. Such capability is vital for optimizing performance across distinguished hardware configurations, such as accelerators, memory hierarchies, and cache systems. Despite that ad hoc optimization techniques, such as compressed/approximate computation, have been enabled for compute-/data-intensive computing for improved performance in distinct hardware settings, there lacks an understanding of 1) the rational behind performance improvement; 2) capability of different optimizations; 3) what optimization to respond to specific computational and memory demands. This work proposes a compression-enabled roofline model to facilitate this adaptability with data compression techniques to balance and transform between computational and memory demands. This model enables applications to adjust in response to the specific strengths and limitations of the underlying hardware and system to optimize resource utilization. The effectiveness of this approach is demonstrated with matrix multiplication kernels on different input sizes, with turning on/off various compression techniques, including 1) low-precision floating point; 2) sparse matrix formulation; and 3) compressed arrays with ZFP. By reducing memory transfer volumes and cache misses and increasing data locality and computational intensity through compression, the specific roofline model can transform between compute and memory bounds to align more efficiently with system capabilities. This advancement not only improves overall performance but also maximizes adaptability in diverse computing environments.
Optical ranging is a problem of estimating the round-trip flight time of a phase- or amplitude-modulated optical beam that reflects off of a target. Frequency- modulated, continuous-wave (FMCW) ranging systems obtain this estimate by performing an interferometric measurement between a local frequency- modulated laser beam and a delayed copy returning from the target. The range estimate is formed by mixing the target-return field with the local reference field on a beamsplitter and detecting the resultant beat modulation. In conventional FMCW ranging, the source modulation is linear in instantaneous frequency, the reference-arm field has many more photons than the target-return field, and the time-of-flight estimate is generated by balanced difference- detection of the beamsplitter output, followed by a frequency-domain peak search. This work focused on determining the maximum-likelihood (ML) estimation algorithm when continuous-time photoncounting detectors are used. It is founded on a rigorous statistical characterization of the (random) photoelectron emission times as a function of the incident optical field, including the deleterious effects caused by dark current and dead time. These statistics enable derivation of the Cramér-Rao lower bound (CRB) on the accuracy of FMCW ranging, and derivation of the ML estimator, whose performance approaches this bound at high photon flux. The estimation algorithm was developed, and its optimality properties were shown in simulation. Experimental data show that it performs better than the conventional estimation algorithms used. The demonstrated improvement is a factor of 1.414 over frequency-domainbased estimation. If the target interrogating photons and the local reference field photons are costed equally, the optimal allocation of photons between these two arms is to have them equally distributed. This is different than the state of the art, in which the local field is stronger than the target return. The optimal processing of the photocurrent processes at the outputs of the two detectors is to perform log-matched filtering followed by a summation and peak detection. This implies that neither difference detection, nor Fourier-domain peak detection, which are the staples of the state-of-the-art systems, is optimal when a weak local oscillator is employed.
Polymers containing poly(ethylene oxide) (PEO) demonstrate superior membrane CO 2 /N 2 separation properties owing to their polar ether oxygen groups exhibiting strong affinity toward CO2. Poly(1,3-dioxolane) (PDXL) shows an ether oxygen content higher than PEO and is expected to have higher CO 2 /N 2 solubility selectivity. However, similar to PEO, the high crystallinity of PDXL greatly reduces its gas permeability. Herein, amorphous PDXL-based hyperbranched polymers were synthesized by ring opening of 1,3-dioxolane (DXL) to form poly(1,3-dioxolane) acrylate (DXLAn) followed by photopolymerization. The repeating unit of DXL (n) or branch length was systematically varied from 4 to 12 to yield amorphous polymers. The chemical and physical properties of the obtained polymers (PDXLAn) were thoroughly evaluated and used to interpret pure- and mixed-gas transport characteristics. The polymers exhibit attractive CO 2 /N 2 and CO 2 /CH 4 separation properties. For example, PDXLA8 exhibits a CO 2 permeability of 220 Barrer and CO 2 /N 2 selectivity of 56 at 35 °C, surpassing Robeson’s 2008 upper bound, and it shows robust separation performance when evaluated with simulated flue gas at 60 °C. This study demonstrates that hyperbranched structures are an effective route to construct amorphous yet highly polar polymers and that chain end groups are instrumental in determining the structural and gas transport characteristics.
Niobium films are a key component in modern two-dimensional superconducting qubits, yet their contribution to the total qubit decay rate is not fully understood. The presence of different layers of materials and interfaces makes it difficult to identify the dominant loss channels in present two-dimensional qubit designs. In this paper, we present the study that directly correlates measurements of RF losses in such films to material parameters by investigating a high-power impulse magnetron sputtered (HiPIMS) film atop a three-dimensional niobium superconducting radio frequency (SRF) resonator. By using a 3D SRF structure, we are able to isolate the niobium film loss from other contributions. Our findings indicate that microwave dissipation in the HiPIMS-prepared niobium films, within the quantum regime, resembles that of record-high intrinsic quality factor of bulk niobium SRF cavities, with lifetimes extending into seconds. Microstructure and impurity level of the niobium film do not significantly affect the losses. These results set the scale of microwave losses in niobium films and show that niobium losses do not dominate the observed coherence times in present two-dimensional superconducting qubit designs, instead highlighting the dominant role of the dielectric oxide in limiting the performance. We can also set a bound for when niobium film losses will become a limitation for qubit lifetimes.
We present a technique for enhancing the estimation of quantum state properties by incorporating approximate prior knowledge about the quantum state. This consists in performing randomized measurements on a quantum processor and comparing the results with those obtained from a classical computer that stores an approximation of the quantum state. We provide unbiased estimators for expectation values of multicopy observables and present performance guarantees in terms of variance bounds that depend on the prior knowledge accuracy. We demonstrate the effectiveness of our approach through experimental and numerical examples detecting mixed-state entanglement, and estimating polynomial approximations of the von Neumann entropy and state fidelities. Published by the American Physical Society 2024
We report state estimation is a fundamental task in power systems. Although distribution systems are increasingly equipped with sensing devices and smart meters, measurements are typically reported at different rates and asynchronously; these aspects pose severe strains on workhorse state estimation algorithms, which are designed to process batches of data collected in a synchronous manner from all the measurement units. In this paper, we develop a novel state estimation algorithm to continuously update the estimate of the state based on measurements received in an asynchronous manner from measurement units. The synthesis of the algorithm hinges on a proximal-point type method, implemented in an online fashion, and capable of processing measurements received sequentially from sensors. A performance analysis is presented by providing bounds on the estimation error in terms of the mean and variance that hold at each iteration and asymptotically. The scheme is also compared with a more traditional Weighted Least Squares estimator that compensates for the lack of measurement data by using, as pseudo measurements, the measurement retrieved during a certain time window. Numerical simulations on the IEEE 37-bus feeder corroborate the analytical findings.
This work revisits the theoretical limits of one-degree-of-freedom wave energy converters. This paper considers the floating sphere used in the Ocean Energy Systems Task 10 Wave Energy Converter modeling and verification effort for analysis. Analytical equations are derived to determine bounds on the motion amplitude, time-averaged power, and power-take-off (PTO) force. A unique result was found that shows the time-averaged power absorbed by a wave energy converter can be defined solely by the inertial properties and radiation hydrodynamic coefficients. In addition, a unique expression for the PTO force amplitude was derived that has provided upper and lower bounds when resistive control is used to maximize power generation. For complex conjugate control, this same expression can only provide a lower bound, as there is theoretically no upper bound. These bounds are used to compare the performance of a floating sphere if it were to extract energy using surge or heave motion. The analysis shows that because of the differences in hydrodynamic coefficients of each oscillating mode, there will be different frequency ranges that provide better power capture efficiency. The influence of a motion constraint on power absorption while also utilizing a nonideal power take-off is examined and found to reduce the losses associated with bidirectional energy flow. The expression to calculate the time-averaged power with a nonideal PTO is modified by the mechanical-to-electrical efficiency and the ratio of the PTO spring and damping coefficients. The PTO spring and damping coefficients were separated in the expression, which allows for limits to be set on the possible values of PTO coefficients to ensure a net flow of power to the grid.
DBSCAN is a well-known density-based clustering algorithm to discover arbitrary shape clusters. While conceptually simple in serial, the algorithm is challenging to efficiently parallelize on manycore GPU architectures. Common pitfalls, such as asynchronous range query calls, result in high thread execution divergence in many implementations. In this paper, we propose a new framework for GPU-accelerated DBSCAN, and describe two tree-based algorithms within that framework. Both algorithms fuse the search for neighbors with updating cluster information, but differ in their treatment of dense regions of the data. We show that the time taken to compute clusters is at most twice that of determination of the neighbors. We compare the proposed algorithms with existing CPU and GPU implementations, and demonstrate their competitiveness and performance using a fast traversal structure (bounding volume hierarchy) for low dimensional data. We also show that the memory usage can be reduced by processing object neighbors dynamically without storing them.
A Deeply Embedded Ring Anchor (DERA) system has been developed as a cost-effective solution for mooring arrays of floating offshore wind turbines (FOWTs) to the seabed. The DERA boasts several key features, including its versatility in various soil types, compact size, compatibility with diverse mooring systems, multi-line potential, and robust performance even under unintentional loading conditions. While prior preliminary studies have provided valuable insights into how the DERA can enhance cost-effectiveness by offering a high load capacity, these studies have predominantly focused on optimizing anchor performance under translational horizontal and vertical loading. However, to design the DERA optimally, we must also consider its ability to handle inclined loading conditions in addition to lateral and axial loadings. Due to its shorter length compared to a conventional caisson, the DERA has less resistance to moments, making it more sensitive to horizontal load capacity and the optimal load attachment depth concerning load angle. For this reason, our study introduces an analytical approach to evaluate the effects of inclined loading on anchor performance, utilizing the previously validated upper bound plastic limit analysis (PLA) method. In investigating the optimal load attachment of the DERA, this paper conducts a parametric study to analyze how factors such as load attachment depth, anchor aspect ratio, and load inclination affect the DERA’s load capacity. Our findings indicate that PLA can serve as a valuable analytical tool for assessing the ultimate load capacity of the DERA, particularly under inclined loading conditions.
Understanding the relevant turbulent transport mechanisms at the edge of a high-performance tokamak is essential for predicting and optimizing the H-mode pedestal structure in future burning plasma devices. Global electromagnetic gyrokinetic simulations will be performed based on representative experimental pedestal scenarios in order to clarify which instabilities are most important for each of the particle and heat transport channels. Edge transport modeling will be performed in order to estimate and bound the particle and heat sources—e.g., the ionization density source and the atomic energy loss channels due to ionization, charge exchange, and radiation. Comparisons will be made with data from the DIII-D, JET, C-Mod and NSTX or MAST experiments.
As part of the U.S. DOE EVs@Scale consortium, the NextGen Profiles (NGP) project presents analysis and results from the study of High Power Charging Electric Vehicles and Battery Charging Infrastructure. High Power Charging equipment is capable of recharging electric vehicle traction batteries at power levels of 200KW and above. The intent of the project is to further understand the most recent technological capabilities of the electric mobility industry related to charging performance. The project aims to develop EV, EVSE, and Fleet characterization testing practices and comprehensive analysis with inputs from key industry stakeholders. The results published in this NextGen Profiles project report provide data and insight for use by numerous entities including modeling and simulation organizations, policy makers, fleet planners, industry stakeholders and the general public involved with the development, deployment and operation of electrified transportation technologies. The factors influencing Electric Vehicle (EV) Direct Current Fast Charging (DCFC), including EV battery specifications, temperature effects on lithium-ion battery and power electronics performance, lithium-ion battery SOC bounding and charging station design considerations are specifically investigated to analyze their impacts on charging station operation and recommendations are made to minimize charge station dwell time, reduce charging costs and mitigate electric grid and charge station congestion. Additional high-power charging results are anticipated in future publications in support of the U.S. DOE EVs@Scale consortium NextGen Profiles project.
Optimal regulator problem for linear plant with bounded control variable and quadratic index performance
Parameter optimization using steepest descent approximation for systems subject to worst bounded disturbance in order to provide performance index minimization
Investigation of the problem of replacing a certain time-varying linear system by a stationary one. Several quadratic criteria are proposed to aid in determining suitable candidate systems. One criterion for choosing the matrix B (in the stationary system) is initial-condition dependent, and another bounds the 'worst case' homogeneous system performance. Both of these criteria produce weighted least square fits.
Spectrally efficient modulation, involving the maximization of the transmitted data rate through a specified bandwidth, may aid in alleviating the problem of spectral congestion in the electromagnetic spectrum. Attention is given to pulse-shaping and nonbinary modulation, and particularly to combined amplitude- and phase-modulation. Theoretical performance limits, such as Shannon's bound, are used to assess the effectiveness of the systems described. Consideration is also allotted to spectrally efficient modulation for terrestrial microwave systems, required to provide 1000 voice channels in the 4- and 6-GHz band.