Search NASA⌕ Search

Engineering topics

Du, Qiang

Publications and source records attributed to Du, Qiang.

At least 19 records
All sourcesNASA NTRSDOE OSTI
DOE OSTI2026

Ginzburg--Landau functionals in the large-graph limit

Ginzburg–Landau (GL) functionals on graphs, which are relaxations of graph-cut functionals on graphs, have yielded a variety of insights in image segmentation and graph clustering. In this paper, we study large-graph limits of GL functionals by taking a functional-analytic view of graphs as nonlocal kernels. For a graph Wn with n nodes, the corresponding graph GL functional GL W n ϵ is an energy for functions on Wn. We minimize GL functionals on sequences of growing graphs that converge to functions called graphons. For such sequences of graphs, we show that the graph GL functional Γ-converges to a continuous and nonlocal functional that we call the graphon GL functional. We investigate the sharp-interface limits of the graph GL and graphon GL functionals, and we relate these limits to a nonlocal total-variation (TV) functional. We express the limiting GL functional in terms of Young measures and thereby obtain a probabilistic interpretation of the minimization problem in the large-graph limit. Finally, to develop intuition about graphon GL functionals, we determine the GL minimizer for several example families of graphons.

Zhang, Edith↗
DOE OSTI2025

Pointing stabilization of a 1 Hz high-power laser via machine learning

Abstract High-power lasers are vital for particle acceleration, imaging, fusion and materials processing, requiring precise control and high-energy delivery. Laser plasma accelerators (LPAs) demand laser positional stability at focus to ensure consistent electron beams in applications such as X-ray free-electron lasers and high-energy colliders. Achieving this stability is especially challenging for the low-repetition-rate lasers in current LPAs. We present a machine learning method that predicts and corrects laser pointing instabilities in real-time using a high-frequency pilot beam. By preemptively adjusting a correction mirror, this approach overcomes traditional feedback limits. Demonstrated on the BELLA petawatt laser operating at the terawatt level (30 mJ amplification), our method achieved root mean square pointing stabilization of 0.34 and 0.59 $\unicode{x3bc} \mathrm{rad}$ in the x and y directions, reducing jitter by 65% and 47%, respectively. This is the first successful application of predictive control for shot-to-shot stabilization in low-repetition-rate laser systems, paving the way for full-energy petawatt lasers and transformative advances across science, industry and security.

Amodio, Alessio↗
DOE OSTI2024

Near-complete extraction of maximum stored energy from large-core fibers using coherent pulse stacking amplification of femtosecond pulses

High field science relies on ultrashort pulse lasers with multi-joule pulse energies for studying light–matter interactions under extreme conditions and for driving particle accelerators and secondary radiation sources of x rays, gamma rays, neutrons, positrons, muons, and protons. Next-generation laser drivers will require a 10 3 -10 4 times increase in pulse repetition rates, producing multi-joule energies at multi-kilowatt average powers to enable practical applications in nuclear engineering, advanced materials, medicine, biology, homeland security, and high-energy physics. Spatially coherently combined femtosecond fiber lasers are recognized as a pathway to these next-generation drivers, with significant practical advantages including high efficiency and the possibility of compact integration. However, chirped pulse amplification in fibers is capable of extracting only a small fraction (usually ~1%) of the maximum stored energy. Here we demonstrate near-complete maximum stored energy extraction with low accumulated nonlinearity from a large-core fiber amplifier using coherent pulse stacking amplification. We have amplified a 81-pulse stacking burst in a 85 µm core chirally coupled core Yb-doped fiber, extracting up to 9.5 mJ (~90% of stored energy) with < 4.5 radians of accumulated nonlinear phase, temporally combined this burst into a single pulse, and achieved 4.2 mJ pulses of 313 fs bandwidth-limited duration after compression. This represents, to our knowledge, the highest energy extracted and compressed into a femtosecond pulse from a single fiber amplifier, enabling approximately two orders of magnitude size reduction of future high-energy coherently spatially combined fiber laser arrays.

47 OTHER INSTRUMENTATION↗
DOE OSTI2024

Generic Multi-Layer Perceptron Inference Accelerator on FPGA (vneuron) v1.0

We have designed and implemented a neural network inference compute engine (vneuron) that can be deployed in the fabric of any FPGA without using special hardware accelerator primitive. The "vneuron" is purely written in verilog, and supports scalable neural network structure with fully connected layers and ReLU activation ( Multi-Layer Perceptron architecture) with 16 bits of precision. We have demonstrated it on an Xilinx Artix 7 FPGA for a 16-input, 8-output MLP with 3 layer, 1600 parameters. It takes 40 DSP48E and 40 BRAM18, and takes 131 clock cycles for computing (1048 ns when clocked at 125MHz). We include PyTorch quantization from a given floating point model, and provide behavioral verification simulation in the disclosed software package.

Du, Qiang↗
DOE OSTI2024

USPAS Digital Low-Level RF firmware and software suite (uspas_llrf) v1.0

The USPAS Digital Low-Level RF firmware and software suite was originally developed for the USPAS 2023 LLRF course, for demonstrating the close loop RF control on the BerkeleyLab Marble FPGA carrier and Zest digitizer platform in controlling the RF field for various accelerators. The suite features Register-transfer level (RTL) designs for the board-support layer, digital signal processing and verification test benches, system-on-chip architecture, and Experimental Physics and Industrial Control System (EPICS) support, utilizing a open-source tool chain.

Du, Qiang↗
DOE OSTI2023

High average power ultrafast laser technologies for driving future advanced accelerators

Large scale laser facilities are needed to advance the energy frontier in high energy physics and accelerator physics. Laser plasma accelerators are core to advanced accelerator concepts aimed at reaching TeV electron electron colliders. In these facilities, intense laser pulses drive plasmas and are used to accelerate electrons to high energies in remarkably short distances. A laser plasma accelerator could in principle reach high energies with an accelerating length that is 1000 times shorter than in conventional RF based accelerators. Notionally, laser driven particle beam energies could scale beyond state of the art conventional accelerators. LPAs have produced multi GeV electron beams in about 20 cm with relative energy spread of about 2 percent, supported by highly developed laser technology. This validates key elements of the US DOE strategy for such accelerators to enable future colliders but extending best results to date to a TeV collider will require lasers with higher average power. While the per pulse energies envisioned for laser driven colliders are achievable with current lasers, low laser repetition rates limit potential collider luminosity. Applications will require rates of kHz to tens of kHz at Joules of energy and high efficiency, and a collider would require about 100 such stages, a leap from current Hz class LPAs. This represents a challenging 1000 fold increase in laser repetition rates beyond current state of the art. Here, this whitepaper describes current research and outlook for candidate laser systems as well as the accompanying broadband and high damage threshold optics needed for driving future advanced accelerators.

47 OTHER INSTRUMENTATION↗
DOE OSTI2023

Minimizing Optimal Transport for Functions with Fixed-Size Nodal Sets

Consider the class of zero-mean functions with fixed L ∞ and L 1 norms and exactly N ϵ N nodal points. Which functions f minimize W p (f + ,f – ), the Wasserstein distance between the measures whose densities are the positive and negative parts? We provide a complete solution to this minimization problem on the line and the circle, which provides sharp constants for previously proven “uncertainty principle”-type inequalities, i.e., lower bounds on N • W p (f + ,f – ). We further show that, while such inequalities hold in many metric measure spaces, they are no longer sharp when the non-branching assumption is violated; indeed, for metric star-graphs, the optimal lower bound on W p (f + ,f – ) is not inversely proportional to the size of the nodal set, N. Here, based on similar reductions, we make connections between the analogous problem of minimizing W p (f + ,f – ) for f defined on Ω C R d with an equivalent optimal domain partition problem.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

A fast two-stage algorithm for non-negative matrix factorization in smoothly varying data

This article reports the study of algorithms for non-negative matrix factorization (NMF) in various applications involving smoothly varying data such as time or temperature series diffraction data on a dense grid of points. Utilizing the continual nature of the data, a fast two-stage algorithm is developed for highly efficient and accurate NMF. In the first stage, an alternating non-negative least-squares framework is used in combination with the active set method with a warm-start strategy for the solution of subproblems. In the second stage, an interior point method is adopted to accelerate the local convergence. The convergence of the proposed algorithm is proved. The new algorithm is compared with some existing algorithms in benchmark tests using both real-world data and synthetic data. Furthermore, the results demonstrate the advantage of the algorithm in finding high-precision solutions.

interior point method↗
DOE OSTI2022

Machine Learning Pattern Recognition Algorithm With Applications to Coherent Laser Combination

Herein we analyze a new kind of machine learning algorithm designed to feedback stabilize coherently combined lasers. This algorithm learns differential, rather than absolute, values of action in phase space, in order to facilitate learning on initially unstable systems. Experiments have shown that this approach can control small-scale spatial beam combination with high stability. In this paper we analyze the algorithm's performance and limitations in depth, showing that it can continuously learn during operation in order to track changes. Using simulation, we extend the application to temporal combination, and show that it scales to more complex instances by combining 81 beams.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2020

Numerical methods for nonlocal and fractional models

Partial differential equations (PDEs) are used with huge success to model phenomena across all scientific and engineering disciplines. However, across an equally wide swath, there exist situations in which PDEs fail to adequately model observed phenomena, or are not the best available model for that purpose. On the other hand, in many situations, nonlocal models that account for interaction occurring at a distance have been shown to more faithfully and effectively model observed phenomena that involve possible singularities and other anomalies. Here, we consider a generic nonlocal model, beginning with a short review of its definition, the properties of its solution, its mathematical analysis and of specific concrete examples. We then provide extensive discussions about numerical methods, including finite element, finite difference and spectral methods, for determining approximate solutions of the nonlocal models considered. In that discussion, we pay particular attention to a special class of nonlocal models that are the most widely studied in the literature, namely those involving fractional derivatives. The article ends with brief considerations of several modelling and algorithmic extensions, which serve to show the wide applicability of nonlocal modelling.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2020

A non-cooperative meta-modeling game for automated third-party calibrating, validating and falsifying constitutive laws with parallelized adversarial attacks

The evaluation of constitutive models, especially for high-risk and high-regret engineering applications, requires efficient and rigorous third-party calibration, validation and falsification. While there are numerous efforts to develop paradigms and standard procedures to validate models, difficulties may arise due to the sequential, manual, and often biased nature of the commonly adopted calibration and validation processes, thus slowing down data collections, hampering the progress towards discovering new physics, increasing expenses and possibly leading to misinterpretations of the credibility and application ranges of proposed models. This work attempts to introduce concepts from game theory and machine learning techniques to overcome many of these existing difficulties. Here, we introduce an automated meta-modeling game where two competing AI agents systematically generate experimental data to calibrate a given constitutive model and to explore its weakness such that the experiment design and model robustness can be improved through competitions. The two agents automatically search for the Nash equilibrium of the meta-modeling game in an adversarial reinforcement learning framework without human intervention. In particular, a protagonist agent seeks to find the more effective ways to generate data for model calibrations, while an adversary agent tries to find the most devastating test scenarios that expose the weaknesses of the constitutive model calibrated by the protagonist. By capturing all possible design options of the laboratory experiments into a single decision tree, we recast the design of experiments as a game of combinatorial moves that can be resolved through deep reinforcement learning by the two competing players. Our adversarial framework emulates idealized scientific collaborations and competitions among researchers to achieve a better understanding of the application range of the learned material laws and prevent misinterpretations caused by conventional AI-based third-party validation. Numerical examples are given to demonstrate the wide applicability of the proposed meta-modeling game with adversarial attacks on both human-crafted constitutive models and machine learning models.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2019

Bedrock v1.0

Bedrock is a code set of firmware libraries written in Verilog used by LBNL for accelerator controls and other applications. A set of commonly used building blocks in DSP (Digital Signal Processing) are present such as a CORDIC implementation, CIC filters, etc. Other useful functions include an in-fabric Ethernet core implementation to communicate with FPGAs over the network and support drivers for components such as ADCs and DACs we have used in our designs. The code base is intended to serve as a support library for HDL designs.

Doolittle, Lawrence↗
DOE OSTI2019

Deterministic stabilization of eight-way 2D diffractive beam combining using pattern recognition

We demonstrate a new method for controlling diffractive, high-power beam combination, sensing phase errors by analyzing the intensity pattern of uncombined side beams at the output. A square array of eight beams is combined with <0.3% stability and 84.6% efficiency. As the channel count is increased, so does the usable information, enabling scaling to large channel counts without significant slowing of control loop response time, an advantage over single-input algorithms.

47 OTHER INSTRUMENTATION↗
DOE OSTI2019

Using a machine learning approach to determine the space group of a structure from the atomic pair distribution function

Here a method is presented for predicting the space group of a structure given a calculated or measured atomic pair distribution function (PDF) from that structure. The method utilizes machine learning models trained on more than 100 000 PDFs calculated from structures in the 45 most heavily represented space groups. In particular, a convolutional neural network (CNN) model is presented which yields a promising result in that it correctly identifies the space group among the top-6 estimates 91.9% of the time. The CNN model also successfully identifies space groups for 12 out of 15 experimental PDFs. Interesting aspects of the failed estimates are discussed, which indicate that the CNN is failing in similar ways as conventional indexing algorithms applied to conventional powder diffraction data. This preliminary success of the CNN model shows the possibility of model-independent assessment of PDF data on a wide class of materials.

99 GENERAL AND MISCELLANEOUS↗