SEARCH · Search NASA

Results for “derivative free 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.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 109 records · Page 6

All sources NASA NTRS DOE OSTI

NASA NTRS1988

A variational dynamic programming approach to robot-path planning with a distance-safety criterion

An approach to robot-path planning is developed by considering both the traveling distance and the safety of the robot. A computationally-efficient algorithm is developed to find a near-optimal path with a weighted distance-safety criterion by using a variational calculus and dynamic programming (VCDP) method. The algorithm is readily applicable to any factory environment by representing the free workspace as channels. A method for deriving these channels is also proposed. Although it is developed mainly for two-dimensional problems, this method can be easily extended to a class of three-dimensional problems. Numerical examples are presented to demonstrate the utility and power of this method.

Suh, Suk-Hwan↗

NASA NTRS1975

Optimal rendezvous in the neighborhood of a circular orbit

The minimum velocity change rendezvous solutions, when the motion may be linearized about a circular orbit, fall into two separate regions; the phase-for-free region and the general region. Phase-for-free solutions are derived from the optimum transfer solutions, require the same velocity change expenditure, but may not be unique. Analytic solutions are presented in two of the three subregions. An algorithm is presented for determining the unique solutions in the general region. Various sources of initial conditions are discussed and three examples presented.

Jones, J. B.↗

DOE OSTI2024

Proximal Galerkin: A Structure-Preserving Finite Element Method for Pointwise Bound Constraints

The proximal Galerkin finite element method is a high-order, low iteration complexity, nonlinear numerical method that preserves the geometric and algebraic structure of pointwise bound constraints in infinite-dimensional function spaces. This paper introduces the proximal Galerkin method and applies it to solve free boundary problems, enforce discrete maximum principles, and develop a scalable, mesh-independent algorithm for optimal design with pointwise bound constraints. This paper also introduces the latent variable proximal point (LVPP) algorithm, from which the proximal Galerkin method derives. When analyzing the classical obstacle problem, we discover that the underlying variational inequality can be replaced by a sequence of second-order partial differential equations (PDEs) that are readily discretized and solved with, e.g., the proximal Galerkin method. Throughout this work, we arrive at several contributions that may be of independent interest. These include (1) a semilinear PDE we refer to as the entropic Poisson equation; (2) an algebraic/geometric connection between high-order positivity-preserving discretizations and certain infinite-dimensional Lie groups; and (3) a gradient-based, bound-preserving algorithm for two-field, density-based topology optimization. The complete proximal Galerkin methodology combines ideas from nonlinear programming, functional analysis, tropical algebra, and differential geometry and can potentially lead to new synergies among these areas as well as within variational and numerical analysis. Open-source implementations of our methods accompany this work to facilitate reproduction and broader adoption.

97 MATHEMATICS AND COMPUTING↗

NASA NTRS1976

Optimal rendezvous in the neighborhood of a circular orbit

The minimum velocity-change rendezvous solutions, when the motion may be linearized about a circular orbit, fall into two separate regions; the phase-for-free region and the general region. Phase-for-free solutions are derived from the optimum transfer solutions, require the same velocity-change expenditure, but may not be unique. Analytic solutions are presented in two of the three subregions. An algorithm is presented for determining the unique solutions in the general region. Various sources of initial conditions are discussed and three examples are presented.

Jones, J. B.↗

NASA NTRS1994

Comparison of transform coding methods with an optimal predictor for the data compression of digital elevation models

Statistical encoding techniques enable the reduction of the number of bits required to encode a set of symbols, and are derived from their probabilities. Huffman encoding is an example of statistical encoding that has been used for error-free data compression. The degree of compression given by Huffman encoding in this application can be improved by the use of prediction methods. These replace the set of elevations by a set of corrections that have a more advantageous probability distribution. In particular, the method of Lagrange Multipliers for minimization of the mean square error has been applied to local geometrical predictors. Using this technique, an 8-point predictor achieved about a 7 percent improvement over an existing simple triangular predictor.

Lewis, Michael↗

DOE OSTI2023

Evaluation of Equiatomic CrMnFeCoNiCu System and Subsequent Derivation of a Non-Equiatomic MnFeCoNiCu Alloy

Investigation into non-equiatomic high-entropy alloys has grown in recent years due to questions about the role of entropy stabilization in forming single-phase solid solutions. Non-equiatomic alloys have been shown to retain the outstanding mechanical properties exhibited by their equiatomic counterparts and even improve electrical, thermal, and magnetic properties, albeit with relaxed composition bounds. However, much remains to understand the processing–structure–property relationships in all classes of so-called high-entropy alloys (HEAs). Here, we are motivated by the natural phenomena of crystal growth and equilibrium conditions to introduce a method of HEA development where controlled processing conditions determine the most probable and stable composition. This is demonstrated by cooling an equiatomic CrMnFeCoNiCu alloy from the melt steadily over 3 days (cooling rate ~4 °C/h). The result is an alloy containing large Cr-rich precipitates and an almost Cr-free matrix exhibiting compositions within the MnFeCoNiCu system (with trace amounts of Cr). From this juncture, it is argued that the most stable composition is within the CrMnFeCoNiCu system rather than the CrMnFeCoNi system. With further optimization and evaluation, a unique non-equiatomic alloy, Mn 17 Fe 21 Co 24 Ni 24 Cu 14 , is derived. The alloy solidifies and recrystallizes into a single-phase face-centered cubic (FCC) polycrystal. In addition to possible applications where Invar is currently utilized, this alloy can be used in fundamental studies that contrast its behavior with its equiatomic counterpart and shed light on the development of HEAs.

36 MATERIALS SCIENCE↗

NASA NTRSDate not supplied

GeoXO Ocean Color Instrument (OCX) Spatial and Temporal Coverage Assessment

NOAA’s next generation Geostationary Extended Observations (GeoXO) satellite system will advance Earth observations from geostationary orbit. GeoXO will supply vital information supporting the U.S. weather, ocean, and climate operations. The recommended three-satellite constellation includes spacecrafts at the current GOES-East and GOES-West positions that will carry an imager, lightning mapper, and an ocean color (OCX) instrument. In this poster we describe an assessment of U.S. Exclusive Economic Zone (EEZ) availability for OCX observations, based on EEZ view geometry, solar angle range over the year, sun glint, as well as cloud climatology. The availability for observations for each EEZ region is estimated per day of the year, as function of atmospheric mass factor (AMF) and sun glint. In addition, cloud climatology – based on statistics derived from three years of GOES-16 ABI cloud mask – is considered to determine the fraction of cloud-free OCX observations for each region as function of time of day and season. The results of such assessments can be used to optimize OCX collections and potentially explore regions outside the EEZ.

Boryana Efremova↗

DOE OSTI2021

Efficient ptychographic phase retrieval via a matrix-free Levenberg-Marquardt algorithm

The phase retrieval problem, where one aims to recover a complex-valued image from far-field intensity measurements, is a classic problem encountered in a range of imaging applications. Modern phase retrieval approaches usually rely on gradient descent methods in a nonlinear minimization framework. Calculating closed-form gradients for use in these methods is tedious work, and formulating second order derivatives is even more laborious. Additionally, second order techniques often require the storage and inversion of large matrices of partial derivatives, with memory requirements that can be prohibitive for data-rich imaging modalities. We use a reverse-mode automatic differentiation (AD) framework to implement an efficient matrix-free version of the Levenberg-Marquardt (LM) algorithm, a longstanding method that finds popular use in nonlinear least-square minimization problems but which has seen little use in phase retrieval. Furthermore, we extend the basic LM algorithm so that it can be applied for more general constrained optimization problems (including phase retrieval problems) beyond just the least-square applications. Since we use AD, we only need to specify the physics-based forward model for a specific imaging application; the first and second-order derivative terms are calculated automatically through matrix-vector products, without explicitly forming the large Jacobian or Gauss-Newton matrices typically required for the LM method. We demonstrate that this algorithm can be used to solve both the unconstrained ptychographic object retrieval problem and the constrained “blind” ptychographic object and probe retrieval problems, under the popular Gaussian noise model as well as the Poisson noise model. We compare this algorithm to state-of-the-art first order ptychographic reconstruction methods to demonstrate empirically that this method outperforms best-in-class first-order methods: it provides excellent convergence guarantees with (in many cases) a superlinear rate of convergence, all with a computational cost comparable to, or lower than, the tested first-order algorithms.

36 MATERIALS SCIENCE↗

DOE OSTI2025

Coupling to rotational manifolds to improve gas-phase pump–probe spectroscopic models

The physical picture of gas-phase optical transitions is normally presented as an isolated two-level system balanced by upward and downward processes. Isolated models assume a phenomenological treatment of collisional dephasing but do not strictly account for collisional population exchange with the rotational baths. While this assumption is valid under low-intensity conditions, where excitation is rate-limiting, isolated models can deviate from Beer’s Law at sufficient pressures and monochromatic intensities when both collisional broadening and power broadening are comparable to (or greater than) lifetime broadening, which are not uncommon conditions for cavity enhanced spectroscopies in the mid-IR spectral range. Although this problem has been addressed by rate-equation models for linear absorption measurements, a general treatment for multi-level quantum mechanical models suitable for non-linear absorption measurements (two-photon/two-color/pump–probe) is lacking. Isolated models require physical parameter inputs that disagree with expected values by at least an order of magnitude. These non-physical models undermine the ability to predict non-linear signal strengths under untested conditions and thereby limit the potential to optimize the sensitivity of non-linear spectroscopies and to expand their analytical applications (e.g., new analytes and/or buffer gases, changes in cavity free-spectral-range, changes in intracavity powers or wavelengths, and accurate investigation of physical phenomena). In this study, we derive bath-coupled models for gaseous pump–probe spectroscopy by application of the quantum Lindblad equation and detailed balance. Bath-coupled models are shown to fit data consistently across variations in intensity and agree with all physically expected values.

Cavity ring-down spectroscopy↗

DOE OSTI2026

Optimal binning of correlated measurements

Experimental measurements are commonly represented on a discrete grid, requiring a balance between granularity and statistical noise. Two strategies have traditionally been used to improve such representations: selecting an appropriate bin width to control discretization error and applying kernel-based smoothing to suppress fluctuations. Despite their shared goal, these approaches have largely developed independently, without a unified statistical description of how discretization and correlation jointly determine measurement precision. Here, we extend the discussion of optimal interval averaging to a correlation-aware setting by Gaussian process regression, which explicitly accounts for correlations among neighboring bins. Starting from first principles, we derive the mean-squared error of discretized measurements and obtain closed-form asymptotic expressions for the optimal bin width and correlation length. When recast in reduced variables, the theory reveals distinct universal scaling laws governing the error in the correlation-free and correlation-controlled regimes. Characterized by intrinsically smooth intensity profiles and counting-based statistics, neutron scattering measurements are well suited for demonstrating the enhanced error contraction enabled by inter-bin correlations. We show that such improvement is achievable over the experimentally accessible Q-range and across multiple instruments and material systems. These results show that explicitly accounting for correlations systematically reshapes the limits of precision in discretized, noise-limited measurements. More broadly, the framework provides a transferable statistical foundation for optimizing data representation, inference, and experimental design across the physical and data sciences.

Tung, Chi-Huan [ORNL] (ORCID:0000000221972074)↗

DOE OSTI2021

Gradient-Informed Design Optimization of Select Nuclear Systems

In this work, we present a gradient-informed design optimization of nuclear reactor core components based on neutronics objectives with both continuous and discrete materials. The main argument in favor of using gradient-informed design optimization is that it scales well with increasing dimensionality of the design space. First, a challenge problem with 121 free parameters is solved with a gradient-informed method and then with a genetic algorithm. Then, a challenge problem to optimize the flux profile of a simplified assembly with eight axial zones is solved. Both challenge problems are solved using directly calculated derivatives from Tools for Sensitivity and Uncertainty Analysis Methodology Implementation (TSUNAMI) in the SCALE package. Furthermore, we demonstrate how a discrete optimization problem—selection of materials for 121 voxels—can be lifted into a continuous problem with mixed materials. In the continuous space, adjoint-based gradients are well-defined, and gradient descent is applicable. Then, a forcing function is introduced that with the selection of an appropriately sized hyperparameter can be used to guide the optimized continuous solution back into a discrete solution. This paper presents an account of the challenges that were faced when applying a gradient-informed optimization algorithm using a Monte Carlo calculation to estimate the gradient information and compares a gradient descent optimization method to a genetic algorithm optimization of the same geometry. Overall, this work demonstrates the potential use of adjoint-based gradient calculations in design optimization of nuclear systems.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

DOE OSTI2025

Analytical gradient-based optimization of CALPHAD model parameters

The calibration of CALPHAD (CALculation of PHAse Diagrams) models involves the solution of a very challenging high-dimensional multiobjective optimization problem. Traditional approaches to parameter fitting predominantly rely on gradient-free methods, which while robust, are computationally inefficient and often scale poorly with model complexity. In this work, we introduce and demonstrate a generalizable framework for analytic gradient-based optimization of the parameters of the CALPHAD model enabled by the recently formalized Jansson derivative technique. This method allows for efficient evaluation of gradients of thermodynamic properties at equilibrium with respect to model parameters, even in the presence of arbitrarily complex internal degrees of freedom. Leveraging these semi-analytic gradients, we employ the conjugate gradient (CG) method to optimize thermodynamic model parameters for four binary alloy systems: Cu-Mg, Fe-Ni, Cr-Ni, and Cr-Fe. Across all systems, CG achieves comparable or superior optimality relative to Bayesian ensemble Markov Chain Monte Carlo (MCMC) with improvements in computational efficiency ranging from one to three orders of magnitude. Furthermore, our results establish a new paradigm for CALPHAD assessments in which high fidelity data-rich model calibration becomes tractable using deterministic gradient-informed algorithms.

CALPHAD↗

DOE OSTI2024

How Much Reserve Fuel: Quantifying the Maximal Energy Cost of System Disturbances

Motivated by the design question of additional fuel needed to complete a task in an uncertain environment, this paper introduces metrics to quantify the maximal additional energy used by a control system in the presence of bounded disturbances, compared to a nominal, disturbance-free system. In particular, we consider the task of finite-time stabilization for a linear, time-invariant system. We compare the nominal energy required to achieve this task in the disturbance-free system to the worst-case energy over all feasible disturbances. Solving for the worst-case energy over all disturbances first leads to an optimal control problem with a least-squares solution, and then an infinite-dimensional optimization problem where we derive an upper bound on the solution. The comparison of energies is accomplished using additive and multiplicative metrics, for which we derive bounds. Simulation examples on an ADMIRE fighter jet model demonstrate the practicability of these metrics, and their variation with the distance of the initial condition from the origin and the task completion time.

koopman operator, resilience↗

NASA NTRS2003

Simple control laws for low-thrust orbit transfers

Two methods are presented by which to determine both a thrust direction and when to apply thrust to effect specified changes in any of the orbit elements except for true anomaly, which is assumed free. The central body is assumed to be a point mass, and the initial and final orbits are assumed closed. Thrust, when on, is of a constant value, and specific impulse is constant. The thrust profiles derived from the two methods are not propellant-optimal, but are based firstly on the optimal thrust directions and location on the osculating orbit for changing each of the orbit elements and secondly on the desired changes in the orbit elements. Two examples of transfers are presented, one in semimajor axis and inclination, and one in semimajor axis and eccentricity. The latter compares favourably with a propellant-optimized transfer between the same orbits. The control laws have few input parameters, but can still capture the complexity of a wide variety of orbit transfers.

orbit transfer control algorithm↗

DOE OSTI2021

Comparative analysis of model-free and model-based HVAC control for residential demand response

In this paper, we present a comparative analysis of model-free reinforcement learning (RL) and model predictive control (MPC) approaches for intelligent control of heating, ventilation, and air-conditioning (HVAC). Deep-Q-network (DQN) is used as a candidate for model-free RL algorithm. The two control strategies were developed for residential demand-response (DR) HVAC system. We considered MPC as our golden standard to compare DQN's performance. The question we tried to answer through this work was, What % of MPC's performance can be achieved by model-free RL approach for intelligent HVAC control?. Based on our test result, RL achieved an average of ≈ 62% daily cost saving of MPC. Considering the pure optimization and model-based nature of MPC methods, the RL showed very promising performance. We believe that the interpretations derived from this comparative analysis provide useful insights to choose from various DR approaches and further enhance the performance of the RL-based methods for building energy managements.

Kurte, Kuldeep↗

NASA NTRS1987

A generalized method for optimization of active noise controllers in three-dimensional spaces

In this investigation, the formulation of an optimal active noise controller for harmonic, enclosed sound fields is derived using an indirect boundary element method. The optimal active noise controller is defined as the volume velocity secondary source strengths which minimize the sound pressure level at a number of discrete interior locations. In addition, the formulation of an optimal active noise controller which minimizes the free field power radiated from a generalized, distributed noise source is presented. Representative results are provided for each formulation. For the enclosure problem, results are given for global control, local control, and passive wall treatments using a rectangular cavity. For the free field radiation problem, the effects of secondary source location and quantity on active noise controller performance are investigated. The relative merits of each formulation are discussed.

Mollo, C. G.↗

NASA NTRS1991

Planning and executing motions for multibody systems in free-fall

The purpose of this research is to develop an end-to-end system that can be applied to a multibody system in free-fall to analyze its possible motions, save those motions in a database, and design a controller that can execute those motions. A goal is for the process to be highly automated and involve little human intervention. Ideally, the output of the system would be data and algorithms that could be put in ROM to control the multibody system in free-fall. The research applies to more than just robots in space. It applies to any multibody system in free-fall. Mathematical techniques from nonlinear control theory were used to study the nature of the system dynamics and its possible motions. Optimization techniques were applied to plan motions. Image compression techniques were proposed to compress the precomputed motion data for storage. A linearized controller was derived to control the system while it executes preplanned trajectories.

Cameron, Jonathan M.↗

DOE OSTI2021

Data-driven Optimal Control Strategy for Virtual Synchronous Generator via Deep Reinforcement Learning Approach

This paper aims at developing a data-driven optimal control strategy for virtual synchronous generator (VSG) in the scenario where no expert knowledge or requirement for system model is available. Firstly, the optimal and adaptive control problem for VSG is transformed into a reinforcement learning task. Specifically, the control variables, i.e., virtual inertia and damping factor, are defined as the actions. Meanwhile, the active power output, angular frequency and its derivative are considered as the observations. Moreover, the reward mechanism is designed based on three preset characteristic functions to quantify the control targets: (1) maintaining the deviation of angular frequency within special limits; (2) preserving well-damped oscillations for both the angular frequency and active power output; (3) obtaining slow frequency drop in the transient process. Next, to maximize the cumulative rewards, a decentralized deep policy gradient algorithm, which features model-free and faster convergence, is developed and employed to find the optimal control policy. With this effort, a data-driven adaptive VSG controller can be obtained. By using the proposed controller, the inverter-based distributed generator can adaptively adjust its control variables based on current observations to fulfill the expected targets in model-free fashion. Finally, simulation results validate the feasibility and effectiveness of the proposed approach.

24 POWER TRANSMISSION AND DISTRIBUTION↗