Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “online algorithm tuning”

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 19 records
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

Multi-objective Bayesian active learning for MeV-ultrafast electron diffraction

Ultrafast electron diffraction using MeV energy beams(MeV-UED) has enabled unprecedented scientific opportunities in the study of ultrafast structural dynamics in a variety of gas, liquid and solid state systems. Broad scientific applications usually pose different requirements for electron probe properties. Due to the complex, nonlinear and correlated nature of accelerator systems, electron beam property optimization is a time-taking process and often relies on extensive hand-tuning by experienced human operators. Algorithm based efficient online tuning strategies are highly desired. Here, we demonstrate multi-objective Bayesian active learning for speeding up online beam tuning at the SLAC MeV-UED facility. The multi-objective Bayesian optimization algorithm was used for efficiently searching the parameter space and mapping out the Pareto Fronts which give the trade-offs between key beam properties. Such scheme enables an unprecedented overview of the global behavior of the experimental system and takes a significantly smaller number of measurements compared with traditional methods such as a grid scan. This methodology can be applied in other experimental scenarios that require simultaneously optimizing multiple objectives by explorations in high dimensional, nonlinear and correlated systems.

43 PARTICLE ACCELERATORS↗
DOE OSTI2026

ReMU: regional minimal updating for model-based derivative-free optimization

Derivative-free optimization (DFO) problems are optimization problems where derivative information is unavailable or extremely difficult to obtain. Model-based DFO solvers have been applied extensively in scientific computing. Powell's NEWUOA (2004) [Powell, The NEWUOA software for unconstrained optimization without derivatives, in Large-Scale Nonlinear Optimization, Nonconvex Optimization and its Applications Vol. 83, G. Di Pillo and M. Roma, eds., Springer, 2006, pp. 255–297] and Wild's POUNDerS (2014) [Wild, Solving derivative-free nonlinear least squares problems with POUNDERS, in Advances and Trends in Optimization with Engineering Applications, T. Terlaky, M.F. Anjos, and S. Ahmed, eds., SIAM, 2017, pp. 529–540] explore the numerical power of the minimal norm Hessian (MNH) model for DFO and contributed to the open discussion on building better models with fewer data to achieve faster numerical convergence. Another decade later, we propose the regional minimal updating (ReMU) models, and extend the previous models into a broader class, including the H 2 norm models [Xie and Yuan, Least H 2 norm updating of quadratic interpolation models for derivative-free trust-region algorithms, IMA J. Numer. Anal. 46 (2025), pp. 21–50]. This paper shows motivation behind ReMU models, computational details, theoretical and numerical results on particular extreme points and the barycentre of ReMU's weight coefficient region, and the associated KKT matrix error and distance. Novel metrics, such as the truncated Newton step error, are proposed to numerically understand the new models' properties. A new algorithmic strategy, based on iteratively adjusting the ReMU model type, is also proposed, and shows numerical advantages by combining and switching between the barycentric model and the classic least Frobenius norm model in an online fashion.

derivative-free trust-region methods↗
DOE OSTI2020

Innovations in optimization and control of accelerators using methods of differential geometry and genetic algorithms (Final Report)

Online tuning of particle accelerators is necessary in order to achieve optimal machine performance. However, it is also a major challenge due to the large parameter space which must be searched and the fact that many of the desir- able objectives compete with one another, and so will not reach their optimal values simultaneously. In order to mitigate these issues, we have explored using dimension-reduction techniques to reduce the size of the parameter space which must be searched and multi-objective genetic algorithms to obtain the sets of tuning parameters which provide pareto-optimal values for the objectives. These methods have enabled us to obtain improved values of the vertical emittance at the Cornell Electron Storage Ring (CESR), and to do so with greater control of orbit errors. Algorithmic tuning is a multidisciplinary endeavour, requiring expertise in beam dynamics, diagnostics, control systems and computer science, and thus a key practical problem is to formulate a common language in which experts with different specialties can communicate. We developed a solution to this problem in the form of a generic accelerator software interface that allows for rapid prototyping of optimization and control algorithms. Our interface is built on the Experimental Physics and Industrial Control Systems (EPICS) and makes possible testing control code in simulation before deployment on real accelerators, as well as deployment of third-party optimization code.

43 PARTICLE ACCELERATORS↗
DOE OSTI2024

Application of deep learning methods for beam size control during user operation at the Advanced Light Source

Past research at the Advanced Light Source (ALS) provided a proof-of-principle demonstration that deep learning methods could be effectively employed to compensate for the significant perturbations to the transverse electron beam size induced by user-controlled adjustments of the insertion devices. However, incorporating these methods into the ALS’ daily operations has faced notable challenges. The complexity of the system’s operational requirements and the significant upkeep demands has restricted their sustained application during user operation. Here, we introduce the development of a more robust neural network (NN)-based algorithm that utilizes a novel online fine-tuning approach and its systematic integration into the day-to-day machine operations. Our analysis emphasizes the process of NN model selection, demonstrates the superior performance of the NN-based method over traditional feedback methods, and examines the effectiveness and resilience of the new algorithm during user-operation scenarios. Published by the American Physical Society 2024

43 PARTICLE ACCELERATORS↗
DOE OSTI2024

Deep learning based x-ray spectrometer for high repetition rate characterization of betatron radiation

Betatron radiation produced from a laser-wakefield accelerator is a broadband, hard x-ray (>1 keV) source that has been used in a variety of applications in medicine, engineering, and fundamental science. Further development and optimization of stable, high repetition rate (HRR) (>1 Hz) betatron sources will provide a means to extend their application base to include single-shot dynamical measurements of ultrafast processes or dense materials. Recent advances in laser technology used in such experiments have enabled increases in shot-rate and system stability, providing improved statistical analysis and detailed parameter scans. However, unique challenges exist at high repetition rate, where data throughput and source optimization are now limited by diagnostic acquisition rates and analysis. Here, we present the development of a machine-learning algorithm for the real-time analysis of betatron radiation. We report on the fielding of this deep learning algorithm for online source characterization at the Institut National de la Recherche Scientifique's Advanced Laser Light Source. By fine-tuning an algorithm originally trained on a fully synthetic dataset using a subset of experimental data, the algorithm can predict the betatron critical energy with a percent error of 7.2 % with a reconstruction time of 1.5 ms, providing a valuable tool for real-time, multi-objective optimization at HRR.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2020

Online Modelling and Optimization of Nonlinear Integrable Systems

Nonlinear integrable optics was recently proposed as a design approach to increase the limits on beam brightness and intensity imposed by fast collective instabilities. To study these systems experimentally, a new research electron and proton storage ring, the Integrable Optics Test Accelerator, was constructed and recently commissioned at Fermilab. Beam-based diagnostics and online modelling of nonlinear systems presents unique challenges - in this paper, we report on our efforts to develop optimization methods suited for such lattices. We explore the effectiveness of neural networks as fast online surrogate estimators, and integrate them into a beam-based tuning algorithm. We also develop a method of knob dimensionality reduction and subsequent robust multivariate optimization for maximizing key performance metrics under complicated lattice optics constraints.

43 PARTICLE ACCELERATORS↗
DOE OSTI2024

Online and Offline Identification of False Data Injection Attacks in Battery Sensors Using a Single Particle Model

The cells in battery energy storage systems are monitored, protected, and controlled by battery management systems whose sensors are susceptible to cyberattacks. False data injection attacks (FDIAs) targeting batteries’ voltage sensors affect cell protection functions and the estimation of critical battery states like the state of charge (SoC). Inaccurate SoC estimation could result in battery overcharging and over discharging, which can have disastrous consequences on grid operations. This paper proposes a three-pronged online and offline method to detect, identify, and classify FDIAs corrupting the voltage sensors of a battery stack. To accurately model the dynamics of the series-connected cells a single particle model is used and to estimate the SoC, the unscented Kalman filter is employed. FDIA detection, identification, and classification was accomplished using a tuned cumulative sum (CUSUM) algorithm, which was compared with a baseline method, the chi-squared error detector. Online simulations and offline batch simulations were performed to determine the effectiveness of the proposed approach. Throughout the batch simulations, the CUSUM algorithm detected attacks, with no false positives, in 99.83% of cases, identified the corrupted sensor in 97% of cases, and determined if the attack was positively or negatively biased in 97% of cases.

25 ENERGY STORAGE↗
DOE OSTI2022

Reinforcement learning for online adaptation of model predictive controllers: Application to a selective catalytic reduction unit

Here we present a novel application of reinforcement learning (RL) for online dynamic tuning of model predictive controllers (MPC). Applying a state-action-reward-state-action (SARSA) algorithm for temporal difference learning with a control-specific reward function improves the error tracking performance of a standard MPC formulation. The proposed RL approach is also readily adaptable to other MPCs, or entirely different control approaches. Practical details for the implementation of the RL-MPC algorithm are also presented. The proposed algorithm is applied to a case study of controlling nitrogen oxide (NO x ) emissions in an industrial selective catalytic reduction (SCR) unit, a control problem characterized by significant nonlinearity and time delay. Along with an RL-MPC formulation for NOx control, another MPC is proposed to mitigate ammonia slip and decrease ammonia consumption in the SCR. Results showing the efficacy of the RL-MPC for NO x control through learning and implementation on the nonlinear SCR dynamic model are presented.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2024

Enhancing Distribution System Resilience: A First-Order Meta-RL Algorithm for Critical Load Restoration

The increasing frequency of extreme events and the integration of distributed energy resources (DERs) into modern grids have elevated the need for resilient and efficient critical load restoration strategies in distribution systems. However, the stochastic nature of renewable DERs, limited energy resource availability and the intricate nonlinearities inherent in complex grid control problem make the problem challenging. Although reinforcement learning (RL) and warm-start RL methods have shown promising results, their performance often falls short in rapidly adapting to new, unseen situations and typically requires exhaustive problem-specific tuning. To address these gaps, we propose a First-Order Meta-based RL (FOM-RL) algorithm within an online framework for adaptive and robust critical load restoration. By harnessing local DERs as the enabling technology, FOM-RL allows the RL agent to swiftly adapt to new unseen scenarios by leveraging previously acquired knowledge of different tasks. Experimental results provide evidence that proposed algorithm learns more efficiently and showcases generalization capabilities across diverse set of operational scenarios. Moreover, a rigorous theoretical analysis yields a tight sublinear regret bound, sensitive to temporal variability, with a task-averaged optimality gap bounded by O(VM+D*/(Tsquare root(M))). These results suggest that optimality improves with task similarity and an increased number of tasks M, reaffirming the efficacy and scalability of the proposed approach in addressing the complexities of critical load restoration in distribution systems.

complexity theory↗
DOE OSTI2026

Microsecond-latency feedback at a particle accelerator by online reinforcement learning on hardware

The commissioning and operation of future large-scale scientific experiments will challenge current tuning and control methods. Reinforcement learning (RL) algorithms are a promising solution due to their ability to dynamically adapt to changing environments and consider delayed consequences. In many real-world applications, RL policies must produce actions in real time, often within microseconds to milliseconds, imposing significant constraints on system latency and computational overhead that conventional machine learning libraries are not designed to handle. To control phenomena in real time at these timescales, RL needs to be deployed on-the-edge, namely on dedicated hardware located near the system it controls, without relying on a host CPU or cloud-based inference. In this work we present the design and deployment of an experience accumulator system in a particle accelerator. In this system, deep-RL algorithms run using hardware acceleration and act within a few microseconds, enabling the use of RL for control of phenomena like beam instabilities. The training uses the collected data offline to reduce the number of operations carried out on the acceleration hardware. The proposed architecture was tested in real experimental conditions at the Karlsruhe research accelerator, a synchrotron light source, where the system was used to control artificially induced horizontal betatron oscillations in real-time, with a control loop period of just 2.7 μs. The results showed a performance comparable to the commercial feedback system available at the accelerator, demonstrating the viability and potential of this approach. Due to the self-learning and reconfiguration capability of this implementation, a seamless application to other control problems is possible. Applications range from particle accelerators to large-scale research and industrial facilities.

FPGA↗
DOE OSTI2020

A Learning-based Supervisory Control Architecture for Electric Vehicle Charging System Paired with Energy Storages

A Machine Learning-based predictive model is developed for optimal dispatching energy storage system integrated with Electric Vehicle battery charger. The model is effectively trained using transfer learning algorithm and successfully validated via measurement data to achieve high fidelity and accuracy. The study findings provide possibilities for future development of the presented approach to implement computational heavy predictive algorithms with multi-step predictions in emerging modular bidirectional charger topologies to realize vehicle-to-grid (V2G) technology. The online tuning can be explored using the new measurements obtained during the operation, henceforth, enhancing the controller performance for integrating multiple DERs into distribution power networks that shows huge interest for the utilities and regulatory agencies. Multiple operation scenatios have been evaluated and simulation results are discussed to verify intended performance of the proposed control solution.

24 POWER TRANSMISSION AND DISTRIBUTION↗
NASA NTRSDate not supplied

Implementation of a Six Degree of Freedom Precision Lunar Landing Algorithm Using Dual Quaternion Representation

In this study, a powered descent guidance algorithm using a unit dual quaternion represen- tation of the vehicle dynamics is implemented in a high-fidelity simulation and on representative flight hardware. This Dual-Quaternion Guidance (DQG) algorithm is applied to the precision lunar landing problem which levies complex constraints upon the trajectory, including state triggered attitude constraints to enable terrain-relative navigation and hazard detection as well as real-time requirements for landing site re-designation. The investigation explores DQG’s usefulness as a mission design tool as well as a real-time guidance algorithm and defines real-time performance requirements for the hazard detection and avoidance (HDA) re-targeting phase of precision lunar landing. The experiment is presented in two parts. First, DQG is implemented within a high-fidelity Monte Carlo simulation to tune the algorithm’s parameters for the simulated vehicle, to refine the mission design, and to develop guidance update timing requirements to perform the HDA maneuver. DQG generates trajectories online for the divert which are tracked by the vehicle’s inner-loop controllers to the targeted landing site. Second, DQG is run on representative hardware to demonstrate real-time operation through a divert maneuver. These results allow for rapid, flexible, optimal mission design satisfying complex constraints, and for the definition of real-time performance requirements for the HDA operations inherent in precision lunar landing. The HDA divert maneuver is found to require guidance trajectory updates in less than three seconds. DQG is found to be too slow to meet this update timing on the descent and landing computer (DLC) in its current implementation. DQG running on alternative hardware can meet the update rate requirement. Algorithm implementation improvements are also recommended which are expected to speed up computation sufficiently to meet requirements on the DLC.

GN&C↗
DOE OSTI2021

Online Optimization of NSLS-II Dynamic Aperture and Injection Transient (Facility Improvement Project)

The Facility Improvement Project (FIP) “Methods of online optimization of NSLS-II storage ring concurrent with user operations” proposed and approved in 2018 is now complete. The first goal is the online optimization of nonlinear beam dynamics to increase the beam lifetime and injection efficiency. We have developed a model-independent optimization technique using advanced algorithms. Using this technique, we increased the NSLS-II dynamic aperture by 20% and reduced the amplitude-dependent tune shift by a factor of two. We applied sextupole optimization to the new high-chromaticity lattice, which has been developed to improve the beam stability and to increase the single-bunch beam intensity. We were able to double the injection efficiency and exceed 90%. The second goal of this project is to provide the top-off injection with minimized perturbations of the beamline user operations. To achieve the transparent injection, we optimize the matching of four storage ring injection kickers reducing the perturbation of the stored beam orbit. The first application of the online optimization resulted in a reduction of the amplitude of residual beam oscillation by a factor of six, from 1300 μm down to 200 μm. Further improvement was limited by the timing jitter of the trigger boards. Then, we replaced all trigger boards with a new design and reduced the timing jitter by a factor of five, from 5 ns down to 1 ns. This upgrade resulted in the injection transient reduction from 200 μm to 120 μm. To optimize the full set of kicker parameters, including the trigger timing, amplitude, and pulse width, we upgraded all kicker power supplies with the capability of tunable waveform width. As a result, we have reduced the injection transient to the limit of 60 μm.

43 PARTICLE ACCELERATORS↗
NASA NTRS1994

State-space self-tuner for on-line adaptive control

Dynamic systems, such as flight vehicles, satellites and space stations, operating in real environments, constantly face parameter and/or structural variations owing to nonlinear behavior of actuators, failure of sensors, changes in operating conditions, disturbances acting on the system, etc. In the past three decades, adaptive control has been shown to be effective in dealing with dynamic systems in the presence of parameter uncertainties, structural perturbations, random disturbances and environmental variations. Among the existing adaptive control methodologies, the state-space self-tuning control methods, initially proposed by us, are shown to be effective in designing advanced adaptive controllers for multivariable systems. In our approaches, we have embedded the standard Kalman state-estimation algorithm into an online parameter estimation algorithm. Thus, the advanced state-feedback controllers can be easily established for digital adaptive control of continuous-time stochastic multivariable systems. A state-space self-tuner for a general multivariable stochastic system has been developed and successfully applied to the space station for on-line adaptive control. Also, a technique for multistage design of an optimal momentum management controller for the space station has been developed and reported in. Moreover, we have successfully developed various digital redesign techniques which can convert a continuous-time controller to an equivalent digital controller. As a result, the expensive and unreliable continuous-time controller can be implemented using low-cost and high performance microprocessors. Recently, we have developed a new hybrid state-space self tuner using a new dual-rate sampling scheme for on-line adaptive control of continuous-time uncertain systems.

Shieh, L. S.↗
NASA NTRS1985

The experimental results of a self tuning adaptive controller using online frequency identification

A fourth-order laboratory dynamic system featuring very low structural damping and a noncolocated actuator-sensor pair has been used to test a novel real-time adaptive controller, implemented in a minicomputer, which consists of a state estimator, a set of state feedback gains, and a frequency-locked loop for real-time parameter identification. The adaptation algorithm employed can correct controller error and stabilize the system for more than 50 percent variation in the plant's natural frequency, compared with a 10 percent stability margin in frequency variation for a fixed gain controller having the same performance as the nominal plant condition. The very rapid convergence achievable by this adaptive system is demonstrated experimentally, and proven with simple, root-locus methods.

Chiang, W.-W.↗
DOE OSTI2024

Unsupervised learning-enabled pulsed infrared thermographic microscopy of subsurface defects in stainless steel

Metallic structures produced with laser powder bed fusion (LPBF) additive manufacturing method (AM) frequently contain microscopic porosity defects, with typical approximate size distribution from one to 100 microns. Presence of such defects could lead to premature failure of the structure. In principle, structural integrity assessment of LPBF metals can be accomplished with nondestructive evaluation (NDE). Pulsed infrared thermography (PIT) is a non-contact, one-sided NDE method that allows for imaging of internal defects in arbitrary size and shape metallic structures using heat transfer. PIT imaging is performed using compact instrumentation consisting of a flash lamp for deposition of a heat pulse, and a fast frame infrared (IR) camera for measuring surface temperature transients. However, limitations of imaging resolution with PIT include blurring due to heat diffusion, sensitivity limit of the IR camera. We demonstrate enhancement of PIT imaging capability with unsupervised learning (UL), which enables PIT microscopy of subsurface defects in high strength corrosion resistant stainless steel 316 alloy. PIT images were processed with UL spatial–temporal separation-based clustering segmentation (STSCS) algorithm, refined by morphology image processing methods to enhance visibility of defects. The STSCS algorithm starts with wavelet decomposition to spatially de-noise thermograms, followed by UL principal component analysis (PCA), fine-tuning optimization, and neural learning-based independent component analysis (ICA) algorithms to temporally compress de-noised thermograms. The compressed thermograms were further processed with UL-based graph thresholding K-means clustering algorithm for defects segmentation. The STSCS algorithm also includes online learning feature for efficient re-training of the model with new data. For this study, metallic specimens with calibrated microscopic flat bottom hole defects, with diameters in the range from 203 to 76 µm, were produced using electro discharge machining (EDM) drilling. While the raw thermograms do not show any material defects, using STSCS algorithm to process PIT images reveals defects as small as 101 µm in diameter. To the best of our knowledge, this is the smallest reported size of a sub-surface defect in a metal imaged with PIT, which demonstrates the PIT capability of detecting defects in the size range relevant to quality control requirements of LPBF-printed high-strength metals.

36 MATERIALS SCIENCE↗
DOE OSTI2025

String instability mitigation of adaptive cruise control without modifying control laws: trajectory shaper and parameter estimation

Vehicle automation technologies equip vehicles with adaptive cruise control (ACC) systems, which relieve driving fatigue. However, recent studies have shown that the current ACC systems are string-unstable (i.e., exacerbate traffic congestion). To achieve string stability, most existing studies directly modify the control algorithms of ACC systems. Alternatively, this study proposes a trajectory shaper (TS)-based method, which only modifies the trajectory information of the predecessor vehicle, so that the ego vehicle driven by a string-unstable ACC system leverages the modified trajectory information to achieve string stability. To devise the TS-based method, an offline-online parameter estimation method integrating batch optimization and an extended Kalman filter is applied to estimate the parameters of an ACC system. The proposed TS-based method is cost-effective during implementation, as it avoids modifying existing ACC control algorithms (which entails a complex analysis of control systems and parameter tuning). In conclusion, the effectiveness of the proposed TS-based method is validated through extensive numerical experiments.

33 ADVANCED PROPULSION SYSTEMS↗
DOE OSTI2022

Curriculum-based Reinforcement Learning for Distribution System Critical Load Restoration

This paper focuses on the critical load restoration problem in distribution systems following major outages. To provide fast online response and optimal sequential decision-making support, a reinforcement learning (RL) based approach is proposed to optimize the restoration. Due to the complexities stemming from the large policy search space, renewable uncertainty, and nonlinearity in a complex grid control problem, directly applying RL algorithms to train a satisfactory policy requires extensive tuning to be successful. To address this challenge, this paper leverages the curriculum learning (CL) technique to design a training curriculum involving a simpler steppingstone problem that guides the RL agent to learn to solve the original hard problem in a progressive and more effective manner. We demonstrate that compared with direct learning, CL facilitates controller training to achieve better performance. To study realistic scenarios where renewable forecasts used for decision-making are in general imperfect, the experiments compare the trained RL controllers against two model predictive controllers (MPCs) using renewable forecasts with different error levels and observe how these controllers can hedge against the uncertainty. Results show that RL controllers are less susceptible to forecast errors than the baseline MPCs and can provide a more reliable restoration process.

24 POWER TRANSMISSION AND DISTRIBUTION↗