Search NASA⌕ Search

SEARCH · Search NASA

Results for “Resource scheduling”

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 739 records · Page 41

Planning and Scheduling for Fleets of Earth Observing Satellites

We address the problem of scheduling observations for a collection of earth observing satellites. This scheduling task is a difficult optimization problem, potentially involving many satellites, hundreds of requests, constraints on when and how to service each request, and resources such as instruments, recording devices, transmitters, and ground stations. High-fidelity models are required to ensure the validity of schedules; at the same time, the size and complexity of the problem makes it unlikely that systematic optimization search methods will be able to solve them in a reasonable time. This paper presents a constraint-based approach to solving the Earth Observing Satellites (EOS) scheduling problem, and proposes a stochastic heuristic search method for solving it.

Frank, Jeremy↗

Estimating Energy Market Schedules using Historical Price Data

The global climate crisis is expected to reshape the energy generation landscape in the coming decades. Increasing integration of non-dispatchable renewable energy resources into energy infrastructures and markets creates uncertainty as well as new opportunities for flexible energy systems. To conduct proper economic evaluation of flexible energy systems, such as integrated energy systems (IES), advancements in modelling of market interactions, such as bidding, is crucial. This work presents a shortcut algorithm which uses two mixed integer linear programs to compute dispatch schedules (e.g., hourly power production targets) that are constrained by the resource's bid information and characteristics (e.g., minimum up and down times) based on historical locational marginal price (LMP) data. The proposed algorithm is approximately 100 times faster and uses orders of magnitude less data than a full production cost model (PCM). We find the shortcut simulator recapitulates generator dispatch signals for the Prescient PCM with approximately 4% error for the RTS-GMLC test system.

electricity generation↗

An economic assessment of behind-the-meter photovoltaics paired with batteries on the Hawaiian Islands.

Due to natural variability and uncertainty, the ever-increasing penetration of solar generation in Hawaii presents challenges to power grid operators to maintain reliable system operation. Demand response (DR) has the potential to be a cost-effective tool for Hawaii to reach its aggressive renewable energy goals while maintaining the reliability of power grids. The Hawaii Public Utilities Commission has approved the Hawaiian Electric Company's revised portfolio of DR programs. The companies have released a grid services purchase agreement and subscribed an initial tranche of load into their DR programs. This paper presents innovative analytical methods and comprehensive economic assessment for distributed photovoltaics (PV) paired with battery energy storage systems (BESSs) for two new DR programs, including fast frequency response and capacity grid service. Optimal dispatch and sizing methods are proposed for the paired system considering different tariff schedules and PV compensation programs across five islands. It was found that while the best resource configuration and potential economic benefits vary with tariff structure, a BESS paired with PV can be optimally dispatched to generate multiple value streams simultaneously. Compensation from DR programs is an important value stream to help increase the cost-effectiveness of the integrated system.

Battery energy storage system↗

Enabling New Operations Concepts for Lunar and Mars Exploration

The planning and scheduling of human space activities is an expensive and time-consuming task that seldom provides the crew with the control, flexibility, or insight that they need. During the past thirty years, scheduling software has seen only incremental improvements; however, software limitations continue to prevent even evolutionary improvements in the operations concept that is used for human space missions. Space missions are planned on the ground long before they are executed in space, and the crew has little input or influence on the schedule. In recent years the crew has been presented with a job jar of activities that they can do whenever they have time, but the contents of the jar is limited to tasks that do not use scarce shared resources and do not have external timing constraints. Consequently, the crew has no control over the schedule of the majority of their own tasks. As humans venture farther from earth for longer durations, it will become imperative that they have the ability to plan and schedule not only their own activities, but also the unattended activities of the systems, equipment, and robots on the journey with them. Significant software breakthroughs are required to enable the change in the operations concept. The crew does not have the time to build or modify the schedule by hand. They only need to issue a request to schedule a task and the system should automatically do the rest. Of course, the crew should not be required to build the complete schedule. Controllers on the ground should contribute the models and schedules where they have the better knowledge. The system must allow multiple simultaneous users, some on earth and some in space. The Mission Operations Laboratory at NASA's Marshall Space flight Center has been researching and prototyping a modeling schema, scheduling engine, and system architecture that can enable the needed paradigm shift - it can make the crew autonomous. This schema and engine can be the core of a planning and scheduling system that would enable multiple planners, some on the earth and some in space, to build one integrated timeline. Its modeling schema can capture all the task requirements; its scheduling engine can build the schedule automatically, and its architecture can allow those (on earth and in space) with the best knowledge of the tasks to schedule them. This paper describes the enabling technology and proposes an operations concept for astronauts autonomously scheduling their activities and the activities around them.

Jaap, John↗

Capacity and Throughput of Urban Air Mobility Vertiports with a First-Come, First-Served Vertiport Scheduling Algorithm

In this paper, a first-come, first-served vertiport scheduling algorithm for Urban Air Mobility (UAM) was exercised to assess and compare the capacity and throughput of various vertiport configurations. The scheduler models each vertiport by the number of vertipads and parking spaces, and manages reservations on timelines for those vertiport resources, at a level of fidelity suitable for fast-time and system-level analyses of UAM concepts and other airspace studies. The paper defines the theoretical model that can be used to estimate the capacity of various vertiport configurations. The theoretical model provides an understanding of the conditions that can lead to either a parking space-limited or a vertipad-limited vertiport. Examples of potential throughput for some vertiport configurations are provided using both a queueing approach as well as a simulated UAM demand scenario. The study demonstrated that a first-come, first-served scheduling approach can have inefficiencies in the use of the vertiport resources. The inefficiencies can increase as the number of resources increases. Nonetheless, 80% or better peak throughput to capacity ratio was observed for most vertiport configurations.

UAM↗

Mass and power modeling of communication satellites

Analytic estimating relationships for the mass and power requirements for major satellite subsystems are described. The model for each subsystem is keyed to the performance drivers and system requirements that influence their selection and use. Guidelines are also given for choosing among alternative technologies which accounts for other significant variables such as cost, risk, schedule, operations, heritage, and life requirements. These models are intended for application to first order systems analyses, where resources do not warrant detailed development of a communications system scenario. Given this ground rule, the models are simplified to 'smoothed' representation of reality. Therefore, the user is cautioned that cost, schedule, and risk may be significantly impacted where interpolations are sufficiently different from existing hardware as to warrant development of new devices.

Price, Kent M.↗

Estimating Energy Market Schedules Using Historical Price Data: Preprint

The global climate crisis is expected to reshape the energy generation landscape in the coming decades. Increasing integration of non-dispatchable renewable energy resources into energy infrastructures and markets increases uncertainty and creates new opportunities for flexible energy systems. To conduct proper economic evaluation of flexible energy systems, such as integrated energy systems (IES), advancements in modelling of market interactions, such as bidding, is crucial. This work presents a shortcut algorithm which uses two mixed integer linear programs to compute dispatch schedules (e.g., hourly power production targets) that are constrained by the resource's bid information and characteristics (e.g., minimum up and down times) based on historical locational marginal price (LMP) data. This is orders of magnitude less data than required for a market clearing calculation with a full production cost model (PCM). We find the shortcut simulator recapitulates generator dispatch signals for the Prescient PCM with approximately 4% error for the RTS-GMLC test system.

electricity generation↗

Utility functions and resource management in an oversubscribed heterogeneous computing environment

We model an oversubscribed heterogeneous computing system where tasks arrive dynamically and a scheduler maps the tasks to machines for execution. The environment and workloads are based on those being investigated by the Extreme Scale Systems Center at Oak Ridge National Laboratory. Utility functions that are designed based on specifications from the system owner and users are used to create a metric for the performance of resource allocation heuristics. Each task has a time-varying utility (importance) that the enterprise will earn based on when the task successfully completes execution. We design multiple heuristics, which include a technique to drop low utility-earning tasks, to maximize the total utility that can be earned by completing tasks. The heuristics are evaluated using simulation experiments with two levels of oversubscription. The results show the benefit of having fast heuristics that account for the importance of a task and the heterogeneity of the environment when making allocation decisions in an oversubscribed environment. Furthermore, the ability to drop low utility-earning tasks allow the heuristics to tolerate the high oversubscription as well as earn significant utility.

97 MATHEMATICS AND COMPUTING↗

Deep Reinforcement Scheduling of Energy Storage Systems for Real-time Voltage Regulation in Unbalanced LV Networks with High PV Penetration

The ever-growing higher penetration of distributed energy resources (DERs) in low-voltage (LV) distribution systems brings both opportunities and challenges to voltage support and regulation. This paper proposes a deep reinforcement learning (DRL)-based scheduling scheme of energy storage systems (ESSs) to mitigate system voltage deviations in unbalanced LV distribution networks. The ESS-based voltage regulation problem is formulated as a multi-stage quadratic stochastic program, with the objective of minimizing the expected total daily voltage regulation cost while satisfying operational constraints. While existing voltage regulation methods are mostly focused on onetime- step control, this paper explores a day-horizon systemwide voltage regulation problem. In other words, the size of action and state spaces are extremely high-dimensional and need to be delicately handled. Furthermore, in order to overcome the difficulty of modeling uncertainties and develop a realtime solution, a learn-to-schedule feedback control framework is proposed by adapting the problem to a model-free DRL setting. The proposed algorithm is tested on a customized 6-bus system and a modified IEEE 34-bus system. Simulation results validate the effectiveness and near-optimality of voltage regulation by ESS in comparison with a deterministic quadratic program solution.

Wang, Shengyi↗

The ground processing simulator - A tool for mission model analysis and planning from a launch site perspective

The Ground Processing Simulator (GPS) is a computer-assisted planning tool designed and developed for Space Shuttle launch site application. Utilizing two programming languages, General Purpose Simulation System and FORTRAN, GPS provides the capability to analyze proposed Shuttle mission models via computer simulation. NASA-developed mission models which specify Shuttle launch rates, mission durations, cargo elements, and designated launch site are tested for feasibility by the simulator. GPS produces facility utilization schedules (including the identification of conflicts), launch data options, cargo element requirement dates, ground support equipment inventory requirements, and other data necessary to assess both the programmatic and launch site resources required to support proposed mission models. The purpose of this computer-assisted analysis is to determine methods which will permit the launching of the maximum number of cargoes per year on schedule, and in the sequence desired, with the minimum expenditure of resources.

Ralph, J. A.↗

A Tariff for Reactive Power - IEEE

This paper describes a suggested tariff or payment for the local supply of reactive power from distributed energy resources. The authors consider four sample customers, and estimate the cost of supply of reactive power for each customer. The power system savings from the local supply of reactive power are also estimated for a hypothetical circuit. It is found that reactive power for local voltage regulation could be supplied to the distribution system economically by customers when new inverters are installed. The inverter would be supplied with a power factor of 0.8, and would be capable of local voltage regulation to a schedule supplied by the utility. Inverters are now installed with photovoltaic systems, fuel cells and microturbines, and adjustable-speed motor drives.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Operator strategies under varying conditions of workload

An attempt was made to operationally define and measure strategic behavior in a complex multiple task environment. The Multi-Attribute Task battery was developed to simulate various aspects of flight and consisted of an auditory communication task, monitoring tasks, a tracking tasks, a resource management task which allowed a wide range of responding patterns, and a scheduling window which allowed operators to predict changes in workload. This battery was validated for its sensitivity to strategic behavior, and baseline measures for each individual task were collected. Twenty-four undergraduate and graduate students then performed the battery for four 64 minute sessions which took place over a period of 2 days. Each subject performed the task battery under four levels of workload, which were presented for equal lengths of time during all four sessions. Results indicated that in general, performance improves as a function of experience with the battery, but that performance decreased as workload level increased. The data also showed that subjects developed strategies for responding to the resource management task which allowed them to manage the high workload levels more efficiently. This particular strategy developed over time but was also associated with errors of complacency. These results are presented along with implications for the aviation field and areas of future research.

Arnegard, Ruth J.↗

MARBLE: A Multi-GPU Aware Job Scheduler for Deep Learning on HPC Systems

Deep learning (DL) has become a key tool for solving complex scientific problems. However, managing the multi-dimensional large-scale data associated with DL, especially atop extant multiple graphics processing units (GPUs) in modern supercomputers poses significant challenges. Moreover, the latest high-performance computing (HPC) architectures bring different performance trends in training throughput compared to the existing studies. Existing DL optimizations such as larger batch size and GPU locality-aware scheduling have little effect on improving DL training throughput performance due to fast CPU-to-GPU connections. Additionally, DL training on multiple GPUs scales sublinearly. Thus, simply adding more GPUs to a system is ineffective. To this end, we design MARBLE, a first-of-its-kind job scheduler, which considers the non-linear scalability of GPUs at the intra-node level to schedule an appropriate number of GPUs per node for a job. By sharing the GPU resources on a node with multiple DL jobs, MARBLE avoids low GPU utilization in current multi-GPU DL training on HPC systems. Our comprehensive evaluation in the Summit supercomputer shows that MARBLE is able to improve DL training performance by up to 48.3% compared to the popular Platform Load Sharing Facility (LSF) scheduler. Compared to the state-of-the-art of DL scheduler, Optimus, MARBLE reduces the job completion time by up to 47%.

Han, Jingoo↗

Modular performance prediction for scientific workflows using Machine Learning

Scientific workflows provide an opportunity for declarative computational experiment design in an intuitive and efficient way. A distributed workflow is typically executed on a variety of resources, and it uses a variety of computational algorithms or tools to achieve the desired outcomes. Such a variety imposes additional complexity in scheduling these workflows on large scale computers. As computation becomes more distributed, insights into expected workload that a workflow presents become critical for effective resource allocation. In this paper, we present a modular framework that leverages Machine Learning for creating precise performance predictions of a workflow. The central idea is to partition a workflow in such a way that makes the task of forecasting each atomic unit manageable and gives us a way to combine the individual predictions efficiently. We recognize a combination of an executable and a specific physical resource as a single module. This gives us a handle to characterize workload and machine power as a single unit of prediction. Overall, our modular technique of creating atomic modules and deployment of longest-path approach to estimate workflow performance, allows the framework to adapt to highly complex nested directed acyclic workflows and scale to new scenarios, since it does not make assumptions of underlying workflow structure. We present performance estimation results of independent workflow modules executed on the XSEDE SDSC Comet cluster using various Machine Learning algorithms. The results provide insights into the behavior and effectiveness of different algorithms in the context of scientific workflow performance prediction.

97 MATHEMATICS AND COMPUTING↗

AI4IO: A suite of AI-based tools for IO-aware scheduling

Traditional workload managers do not have the capacity to consider how IO contention can increase job runtime and even cause entire resource allocations to be wasted. Whether from bursts of IO demand or parallel file systems (PFS) performance degradation, IO contention must be identified and addressed to ensure maximum performance. In this paper, we present AI4IO (AI for IO), a suite of tools using AI methods to prevent and mitigate performance losses due to IO contention. AI4IO enables existing workload managers to become IO-aware. Currently, AI4IO consists of two tools: PRIONN and CanarIO. PRIONN predicts IO contention and empowers schedulers to prevent it. CanarIO mitigates the impact of IO contention when it does occur. We measure the effectiveness of AI4IO when integrated into Flux, a next-generation scheduler, for both small- and large-scale IO-intensive job workloads. Our results show that integrating AI4IO into Flux improves the workload makespan up to 6.4%, which can account for more than 18,000 node-h of saved resources per week on a production cluster in our large-scale workload.

Wyatt, II, Michael R.↗

Applying an improved rapid impact assessment matrix method to strategic environmental assessment of urban planning in China

Strategic environmental assessment (SEA) has become an increasingly important decision-support tool for providing information on the environmental implications of a policy, plan, or program. The goal is to safeguard the environment and promote sustainable development at the strategic level. Despite major progress in implementing SEA practices internationally, developing countries, such as China, often lag behind in applying SEA methodology. Lack of available data and time constraints arising from tight schedules create problems. The rapid impact assessment matrix (RIAM) is a potential resource for breaking through such difficulties. Our analysis of RIAM applications suggested that it could become a tool for evaluating strategic alternatives because of its applicability in interdisciplinary settings, its transparency, and its short implementation timeframe. To make it more suitable for the SEA context, we have developed two major improvements to the conventional RIAM process: assignment of weights to assessment indicators and the development of an integrated environmental assessment score (IES). The improved RIAM process was employed in an SEA of the development plan for the Nansha District in Guangzhou, the capital city of Guangdong Province in China. It was used to assess five alternatives for development in Wanqingsha (WQS), a subunit of Nansha, where important ecological resources are located and where industrial development could impact the air quality in the neighboring Hong Kong Special Administrative Region (HKSAR). The evaluation identified WQS-A04 as the preferred alternative. This alternative involved a minimal amount of industrial development – 10% compared with the most intense development alternative – and included important wetland preservation plans. The assessment results have been incorporated into the officially approved development plan for Nansha. The improved RIAM methodology is well adapted to the technical aims of SEA and decision-making structures in China. It offers the potential for delivering timely and quality results to decision-making processes. To achieve the desired efficiency, it is recommended that an SEA procedure take into account findings acquired from an improved RIAM application at an early stage, then, at a later stage, results from more comprehensive assessments conducted using more sophisticated methods should be added, if time and data are available. - Highlights: • RIAM is a potential SEA tool due to its multidisciplinary setting and rapid process. • Weighting index and computing IES are raised for improving RIAM when applied to SEAs. • The improved RIAM boosts SEA effectiveness by timely responding to decision-making. • The improved RIAM is more suitable to an SEA bothered by tight time or data shortage. • The improved RIAM adapts to the technical aims of SEA and decision features in China.

54 ENVIRONMENTAL SCIENCES↗

Robust optimisation-based microgrid scheduling with islanding constraints

This paper proposes a robust optimization based optimal scheduling model for microgrid operation considering constraints of islanding capability. Our objective is to minimize the total operation cost, including generation cost and spinning reserve cost of local resources as well as purchasing cost of energy from the main grid. In order to ensure the resiliency of a microgrid and improve the reliability of the local electricity supply, the microgrid is required to maintain enough spinning reserve (both up and down) to meet local demand and accommodate local renewable generation when the supply of power from the main grid is interrupted suddenly, i.e., microgrid transitions from grid-connected into islanded mode. Prevailing operational uncertainties in renewable energy resources and load are considered and captured using a robust optimization method. With proper robust level, the solution of the proposed scheduling model ensures successful islanding of the microgrid with minimum load curtailment and guarantees robustness against all possible realizations of the modeled operational uncertainties. Numerical simulations on a microgrid consisting of a wind turbine, a PV panel, a fuel cell, a micro-turbine, a diesel generator and a battery demonstrate the effectiveness of the proposed scheduling model.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗

Satellite Power Systems (SPS) concept definition study, exhibit C. Volume 2, part 2: System engineering, cost and programmatics

Volume 2, Part 2, of a seven volume Satellite Power Systems (SPS) report is presented. Part 2 covers cost and programmatics and is divided into four sections. The first section gives illustrations of the SPS reference satellite and rectenna concept, and an overall scenario for SPS space transportation involvement. The second section presents SPS program plans for the implementation of PHASE C/D activities. These plans describe SPS program schedules and networks, critical items of systems evolution/technology development, and the natural resources analysis. The fourth section presents summary comments on the methods and rationale followed in arriving at the results documented. Suggestions are also provided in those areas where further analysis or evaluation will enhance SPS cost and programmatic definitions.

Hanley, G. M.↗