Search NASA⌕ Search

SEARCH · Search NASA

Results for “greedy algorithm”

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.

82 records · Page 5
All sourcesNASA NTRSDOE OSTI
DOE OSTI2019

ALTSched: Improved Scheduling for Time-domain Science with LSST

Abstract Telescope scheduling is the task of determining the best sequence of observations (pointings and filter choices) for a survey system. Because it is computationally intractable to optimize over all possible multiyear sequences of observations, schedulers use heuristics to pick the best observation at a given time. A greedy scheduler selects the next observation by choosing whichever one maximizes a scalar merit function, which serves as a proxy for the scientific goals of the telescope. This sort of bottom-up approach for scheduling is not guaranteed to produce a schedule for which the sum of merit over all observations is maximized. As an alternative to greedy schedulers, we introduce ALTS ched, which takes a top-down approach to scheduling. Instead of considering only the next observation, ALTS ched makes global decisions about which area of sky and which filter to observe in, and then refines these decisions into a sequence of observations taken along the meridian to minimize airmass. We implement ALTS ched for the Large Synoptic Survey Telescope (LSST) and show that it equals or outperforms the baseline greedy scheduler in essentially all quantitative performance metrics. Due to its simplicity, our implementation is considerably faster than O p S im , the simulated greedy scheduler currently used by the LSST Project: a full 10-yr survey can be simulated in 4 minutes, as opposed to tens of hours for O p S im . LSST’s hardware is fixed, so improving the scheduling algorithm is one of the only remaining ways to optimize LSST’s performance. We see ALTS ched as a prototype scheduler that gives a lower bound on the performance achievable by LSST.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2025

FIRE: A Failure-Adaptive RL Framework for Edge Computing Migrations

In edge computing, users' service profiles are migrated between edge servers due to user mobility. Reinforcement Learning (RL) frameworks have been proposed to do so, often trained on simulated data. However, existing RL frameworks overlook occasional server failures, which although rare, impact latency-sensitive applications like AR/VR and real- time obstacle detection. These rare failures, being not adequately represented in historical training data, pose a challenge for data-driven RL algorithms. We introduce FIRE, a framework that adapts to rare events by training a RL policy in an edge computing digital twin environment. We propose FIRE-ImRE, an importance sampling-based Q-learning algorithm, which samples rare events proportionally to their impact on the value function. FIRE considers delay, migration, failure, and backup placement costs across individual and shared service profiles. We prove FIRE-ImRE's boundedness and convergence to optimality. Next, we introduce novel deep Q-learning (FIRE-ImDQL) and actor critic (FIRE-ImACRE) versions of our algorithm to enhance scalability. Here, we extend our framework to accommodate users with varying risk tolerances of rare failure events. Through trace-driven experiments, we show that FIRE reduces edge computing costs compared to vanilla RL and the greedy baseline in the event of failures.

Edge computing↗
DOE OSTI2024

Improved stellarator permanent magnet designs through combined discrete and continuous optimizations

A common optimization problem in the areas of magnetized plasmas and fusion energy is the design of magnets to produce a given three-dimensional magnetic field distribution to high precision. When designing arrays of permanent magnets for stellarator plasma confinement, such problems have tens of thousands of degrees of freedom whose solutions, for practical reasons, should be constrained to discrete spaces. We perform a direct comparison between two algorithms that have been developed previously for this purpose, and demonstrate that composite procedures that apply both algorithms in sequence can produce substantially improved results. One approach uses a continuous, quasi-Newton procedure to optimize the dipole moments of a set of magnets and then projects the solution onto a discrete space. The second uses an inherently discrete greedy optimization procedure that has been enhanced and generalized for this work. Further, the approaches are both applied to design arrays cubic rare-Earth permanent magnets to confine a quasi-axisymmetric plasma with a magnetic field on axis of 0.5 T. The first approach tends to find solutions with higher field accuracy, whereas the second can find solutions with substantially (up to 30%) fewer magnets. When the approaches are combined, they can obtain solutions with magnet quantities comparable to the second approach while matching the field accuracy of the first.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2021

Adaptive resource allocation for surrogate modeling of systems comprised of multiple disciplines with varying fidelity

We present an adaptive algorithm for constructing surrogate models for integrated systems composed of a set of coupled components. With this goal we introduce ‘coupling’ variables with a priori unknown distributions that allow approximations of each component to be built independently. Once built, the surrogates of the components are combined and used to predict system-level quantities of interest (QoI) at a fraction of the cost of interrogating the full system model. We use a greedy experimental design procedure, based upon a modification of Multi-Index Stochastic Collocation (MISC), to minimize the error of the combined surrogate. This is achieved by refining each component surrogate in accordance with its relative contribution to error in the approximation of the system-level QoI. Our adaptation of MISC is a multi-fidelity procedure that can leverage ensembles of models of varying cost and accuracy, for one or more components, to produce estimates of system-level QoI. Several numerical examples demonstrate the efficacy of the proposed approach on systems involving feed-forward and feedback coupling. For a fixed computational budget, the proposed algorithm is able to produce approximations that are orders of magnitude more accurate than approximations that treat the integrated system as a black-box.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

Bayesian sequential optimal experimental design for nonlinear models using policy gradient reinforcement learning

We present a mathematical framework and computational methods for optimally designing a finite sequence of experiments. This sequential optimal experimental design (sOED) problem is formulated as a finite-horizon partially observable Markov decision process (POMDP) under a Bayesian setting and with information-theoretic utilities. The formulation is general and may accommodate continuous random variables, non-Gaussian posteriors, and nonlinear forward models. The sOED design policy incorporates elements of feedback and lookahead simultaneously, and we show it to generalize the commonly-used batch and greedy design strategies. We solve for the sOED policy using the policy gradient (PG) method from reinforcement learning, and provide a derivation for the PG expression in the sOED context. Adopting an actor-critic approach, the policy and value functions are parameterized using deep neural networks and improved via PG estimates produced from simulated episodes of designs and observations. The new PG-sOED algorithm is first validated on a linear-Gaussian benchmark, and then compared against other design baselines on a sensor movement problem for contaminant source inversion in a convection-diffusion field. As a result, we provide explanation for the policy behaviors using knowledge of the underlying physical process.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2020

Weighted greedy-optimal design of computer experiments for kernel-based and Gaussian process model emulation and calibration

This article is concerned with the approximation of high-dimensional functions by kernel-based methods. Motivated by uncertainty quantification, which often necessitates the construction of approximations that are accurate with respect to a probability density function of random variables, we aim at minimizing the approximation error with respect to a weighted $L^p$-norm. We present a greedy procedure for designing computer experiments based upon a weighted modification of the pivoted Cholesky factorization. The method successively generates nested samples with the goal of minimizing error in regions of high probability. Numerical experiments validate that this new importance sampling strategy is superior to other sampling approaches, especially when used with non-product probability density functions. We also show how to use the proposed algorithm to efficiently generate surrogates for inferring unknown model parameters from data.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2017

Bayesian K-SVD Using Fast Variational Inference

Recent work in signal processing in general and image processing in particular deals with sparse representation related problems. Two such problems are of paramount importance: an overriding need for designing a well-suited overcomplete dictionary containing a redundant set of atoms— i.e., basis signals—and how to find a sparse representation of a given signal with respect to the chosen dictionary. Dictionary learning techniques, among which we find the popular K-singular value decomposition algorithm, tackle these problems by adapting a dictionary to a set of training data. A common drawback of such techniques is the need for parameter-tuning. In order to overcome this limitation, we propose a fullyautomated Bayesian method that considers the uncertainty of the estimates and produces a sparse representation of the data without prior information on the number of non-zeros in each representation vector. We follow a Bayesian approach that uses a three-tiered hierarchical prior to enforce sparsity on the representations and develop an efficient variational inference framework that reduces computational complexity. Furthermore, we describe a greedy approach that speeds up the whole process. Lastly, we present experimental results that show superior performance on two different applications with real images: denoising and inpainting.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2017

A Real-Time Greedy-Index Dispatching Policy for using PEVs to Provide Frequency Regulation Service

This article presents a real-time greedy-index dispatching policy (GIDP) for using plug-in electric vehicles (PEVs) to provide frequency regulation services. A new service cost allocation mechanism is proposed to award PEVs based on the amount of service they provided, while considering compensations for delayed-charging and reduction of battery lifetime due to participation of the service. The GIDP transforms the optimal dispatch problem from a high-dimensional space into a one-dimensional space while preserving the solution optimality. When solving the transformed problem in real-time, the global optimality of the GIDP solution can be guaranteed by mathematically proved “indexability”. Because the GIDP index can be calculated upon the PEV’s arrival and used for the entire decision making process till its departure, the computational burden is minimized and the complexity of the aggregator dispatch process is significantly reduced. Finally, simulation results are used to evaluate the proposed GIDP, and to demonstrate the potential profitability from providing frequency regulation service by using PEVs.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗
DOE OSTI2016

A comprehensive formulation for volumetric modulated arc therapy planning

Purpose: Volumetric modulated arc therapy (VMAT) is a widely employed radiation therapy technique, showing comparable dosimetry to static beam intensity modulated radiation therapy (IMRT) with reduced monitor units and treatment time. However, the current VMAT optimization has various greedy heuristics employed for an empirical solution, which jeopardizes plan consistency and quality. The authors introduce a novel direct aperture optimization method for VMAT to overcome these limitations. Methods: The comprehensive VMAT (comVMAT) planning was formulated as an optimization problem with an L2-norm fidelity term to penalize the difference between the optimized dose and the prescribed dose, as well as an anisotropic total variation term to promote piecewise continuity in the fluence maps, preparing it for direct aperture optimization. A level set function was used to describe the aperture shapes and the difference between aperture shapes at adjacent angles was penalized to control MLC motion range. A proximal-class optimization solver was adopted to solve the large scale optimization problem, and an alternating optimization strategy was implemented to solve the fluence intensity and aperture shapes simultaneously. Single arc comVMAT plans, utilizing 180 beams with 2° angular resolution, were generated for a glioblastoma multiforme case, a lung (LNG) case, and two head and neck cases—one with three PTVs (H&N{sub 3PTV}) and one with foue PTVs (H&N{sub 4PTV})—to test the efficacy. The plans were optimized using an alternating optimization strategy. The plans were compared against the clinical VMAT (clnVMAT) plans utilizing two overlapping coplanar arcs for treatment. Results: The optimization of the comVMAT plans had converged within 600 iterations of the block minimization algorithm. comVMAT plans were able to consistently reduce the dose to all organs-at-risk (OARs) as compared to the clnVMAT plans. On average, comVMAT plans reduced the max and mean OAR dose by 6.59% and 7.45%, respectively, of the prescription dose. Reductions in max dose and mean dose were as high as 14.5 Gy in the LNG case and 15.3 Gy in the H&N{sub 3PTV} case. PTV coverages measured by D95, D98, and D99 were within 0.25% of the prescription dose. By comprehensively optimizing all beams, the comVMAT optimizer gained the freedom to allow some selected beams to deliver higher intensities, yielding a dose distribution that resembles a static beam IMRT plan with beam orientation optimization. Conclusions: The novel nongreedy VMAT approach simultaneously optimizes all beams in an arc and then directly generates deliverable apertures. The single arc VMAT approach thus fully utilizes the digital Linac’s capability in dose rate and gantry rotation speed modulation. In practice, the new single VMAT algorithm generates plans superior to existing VMAT algorithms utilizing two arcs.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2022

Fast yaw optimization for wind plant wake steering using Boolean yaw angles

Abstract. In wind plants, turbines can be yawed into the wind to steer their wakes away from downstream turbines and achieve an overall increase in plant power. Mathematical optimization is typically used to determine the best yaw angles at which to operate the turbines in a plant. In this paper, we present a new heuristic to rapidly determine the yaw angles in a wind plant. In this method, we define the turbine yaw angles as Boolean – either yawed at a predefined angle or nonyawed – as opposed to the typical methods of defining yaw angles as continuous or with fine discretizations. We then optimize which turbines should be yawed with an algorithm that sweeps through the turbines from the most upstream to the most downstream. We demonstrate that our new Boolean optimization method can find turbine yaw angles that perform well compared to a traditionally used gradient-based optimizer for which the yaw angles are defined as continuous. There is less than 0.6 % difference in the optimized power between the two optimization methods for randomly placed turbine layouts and less than a 0.6 % difference in the optimal annual energy production between the two optimization methods for a real wind farm. Additionally, we show that our new method is much more computationally efficient than the traditional method. For plants with nonzero optimal yaw angles, our new method is generally able to solve for the turbine yaw angles 50–150 times faster, and in some extreme cases up to 500 times faster, than the traditional method.

17 WIND ENERGY↗