Search NASA⌕ Search

SEARCH · Search NASA

Results for “allocation methods”

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 289 records · Page 16

Stochastic pre-event preparation for enhancing resilience of distribution systems

Extreme weather events are the common causes for power supply interruptions and power outages in electrical distribution systems. Improving the distribution system and enhancing its resilience is becoming crucial due to the increased frequency of extreme weather events. Preparation and allocation of multiple flexible resources, such as mobile resources, fuel resources, and labor resources before extreme weather events can mitigate the effects of extreme weather events and enhance the resilience of power distribution systems. Here, in this paper, a two-stage stochastic mixed-integer linear programming (SMILP) is proposed to optimize the preparation and resource allocation process for upcoming extreme weather events, which leads to faster and more efficient post-event restoration. The objective of the proposed two-stage SMILP is to maximize the served load and minimize the operating cost of flexible resources. The first stage in the optimization problem selects the amounts and locations of different resources. The second stage considers the operational constraints of the distribution system and repair crew scheduling constraints. The proposed stochastic pre-event preparation model is solved by a scenario decomposition method, Progressive Hedging (PH), to ease the computational complexity introduced by a large number of scenarios. Furthermore, to show the impact of solar photovoltaic (PV) generation on system resilience, three types of PV systems are considered during a power outage and the resilience improvements with different PV penetration levels are compared. Numerical results from simulations on a large-scale (more than 10,000 nodes) distribution feeder have been used to validate the effectiveness and scalability of the proposed method.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Modelling grain-scattered ultrasound in austenitic stainless-steel welds: A hybrid model

The ultrasonic inspection of austenitic stainless steel welds can be challenging due to their coarse grain structure, charaterised by preferentially oriented, elongated grains. The anisotropy of the weld is manifested as both a ‘steering’ of the beam and the back-scatter of energy due to the macroscopic granular structure of the weld. However, the influence of weld properties, such as mean grain size and orientation distribution, on the magnitude of scattered ultrasound is not well understood. A hybrid model has been developed to allow the study of grain-scatter effects in austenitic welds. An efficient 2D Finite Element (FE) method is used to calculate the complete scattering response from a single elliptical austenitic grain of arbitrary length and width as a function of the specific inspection frequency. A grain allocation model of the weld is presented to approximate the characteristic structures observed in austenitic welds and the complete scattering behaviour of each grain calculated. This model is incorporated into a semi-analytical framework for a single-element inspection of a typical weld in immersion. Experimental validation evidence is demonstrated indicating excellent qualitative agreement of SNR as a function of frequency and a minimum SNR difference of 2 dB at a centre frequency of 2.25 MHz. Additionally, an example Monte-Carlo study is presented detailing the variation of SNR as a function of the anisotropy distribution of the weld, and the application of confidence analysis to inform inspection development.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

WE-AB-209-10: Optimizing the Delivery of Sequential Fluence Maps for Efficient VMAT Delivery

Purpose: To develop an optimization model and solution approach for computing MLC leaf trajectories and dose rates for high quality matching of a set of optimized fluence maps to be delivered sequentially around a patient in a VMAT treatment. Methods: We formulate the fluence map matching problem as a nonlinear optimization problem where time is discretized but dose rates and leaf positions are continuous variables. For a given allotted time, which is allocated across the fluence maps based on the complexity of each fluence map, the optimization problem searches for the best leaf trajectories and dose rates such that the original fluence maps are closely recreated. Constraints include maximum leaf speed, maximum dose rate, and leaf collision avoidance, as well as the constraint that the ending leaf positions for one map are the starting leaf positions for the next map. The resulting model is non-convex but smooth, and therefore we solve it by local searches from a variety of starting positions. We improve solution time by a custom decomposition approach which allows us to decouple the rows of the fluence maps and solve each leaf pair individually. This decomposition also makes the problem easily parallelized. Results: We demonstrate method on a prostate case and a head-and-neck case and show that one can recreate fluence maps to high degree of fidelity in modest total delivery time (minutes). Conclusion: We present a VMAT sequencing method that reproduces optimal fluence maps by searching over a vast number of possible leaf trajectories. By varying the total allotted time given, this approach is the first of its kind to allow users to produce VMAT solutions that span the range of wide-field coarse VMAT deliveries to narrow-field high-MU sliding window-like approaches.

60 APPLIED LIFE SCIENCES↗

Fifteen-Year Radiotherapy Outcomes of the Randomized PORTEC-1 Trial for Endometrial Carcinoma

Purpose: To evaluate the very long-term results of the randomized Post Operative Radiation Therapy in Endometrial Carcinoma (PORTEC)-1 trial for patients with Stage I endometrial carcinoma (EC), focusing on the role of prognostic factors for treatment selection and the long-term risk of second cancers. Patients and Methods: The PORTEC trial (1990-1997) included 714 patients with Stage IC Grade 1-2 or Stage IB Grade 2-3 EC. After surgery, patients were randomly allocated to external-beam pelvic radiotherapy (EBRT) or no additional treatment (NAT). Analysis was by intention to treat. Results: 426 patients were alive at the date of analysis. The median follow-up time was 13.3 years. The 15-year actuarial locoregional recurrence (LRR) rates were 6% for EBRT vs. 15.5% for NAT (p < 0.0001). The 15-year overall survival was 52% vs. 60% (p = 0.14), and the failure-free survival was 50% vs. 54% (p = 0.94). For patients with high-intermediate risk criteria, the 15-year overall survival was 41% vs. 48% (p = 0.51), and the 15-year EC-related death was 14% vs. 13%. Most LRR in the NAT group were vaginal recurrences (11.0% of 15.5%). The 15-year rates of distant metastases were 9% vs. 7% (p = 0.25). Second primary cancers had been diagnosed over 15 years in 19% of all patients, 22% vs. 16% for EBRT vs. NAT (p = 0.10), with observed vs. expected ratios of 1.6 (EBRT) and 1.2 (NAT) compared with a matched population (p = NS). Multivariate analysis confirmed the prognostic significance of Grade 3 for LRR (hazard ratio [HR] 3.4, p = 0.0003) and for EC death (HR 7.3, p < 0.0001), of age >60 (HR 3.9, p = 0.002 for LRR and 2.7, p = 0.01 for EC death) and myometrial invasion >50% (HR 1.9, p = 0.03 and HR 1.9, p = 0.02). Conclusions: The 15-year outcomes of PORTEC-1 confirm the relevance of HIR criteria for treatment selection, and a trend for long-term risk of second cancers. EBRT should be avoided in patients with low- and intermediate-risk EC.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Understanding activity engagement across weekdays and weekend days: A multivariate multiple discrete-continuous modeling approach

This paper is motivated by the increasing recognition that modeling activity-travel demand for a single day of the week, as is done in virtually all travel forecasting models, may be inadequate in capturing underlying processes that govern activity-travel scheduling behavior. The considerable variability in daily travel suggests that there are important complementary relationships and competing tradeoffs involved in scheduling and allocating time to various activities across days of the week. Both limited survey data availability and methodological challenges in modeling week-long activity-travel schedules have precluded the development of multi-day activity-travel demand models. With passive and technology-based data collection methods increasingly in vogue, the collection of multi-day travel data may become increasingly commonplace in the years ahead. This paper addresses the methodological challenge associated with modeling multi-day activity-travel demand by formulating a multivariate multiple discrete-continuous probit (MDCP) model system. The comprehensive framework ties together two MDCP model components, one corresponding to weekday time allocation and the other to weekend activity-time allocation. By tying the two MDCP components together, the model system also captures relationships in activity-time allocation between weekdays on the one hand and weekend days on the other. Model estimation on a week-long travel diary data set from the United Kingdom shows that there are significant inter-relationships between weekdays and weekend days in activity-travel scheduling behavior. In conclusion, the model system presented in this paper may serve as a higher-level multi-day activity scheduler in conjunction with existing daily activity-based travel models.

30 DIRECT ENERGY CONVERSION↗

Attributing land-use change carbon emissions to exported biomass

In this study, a simple, transparent and robust method is developed in which land-use change (LUC) emissions are retrospectively attributed to exported biomass products based on the agricultural area occupied for the production. LUC emissions account for approximately one-fifth of current greenhouse gas emissions. Increasing agricultural exports are becoming an important driver of deforestation. Brazil and Indonesia are used as case studies due to their significant deforestation in recent years. According to our study, in 2007, approximately 32% and 15% of the total agricultural land harvested and LUC emissions in Brazil and Indonesia respectively were due to exports. The most important exported single items with regard to deforestation were palm oil for Indonesia and bovine meat for Brazil. To reduce greenhouse gas (GHG) emissions effectively worldwide, leakage of emissions should be avoided. This can be done, for example, by attributing embodied LUC emissions to exported biomass products. With the approach developed in this study, controversial attribution between direct and indirect LUC and amortization of emissions over the product life cycle can be overcome, as the method operates on an average basis and annual level. The approach could be considered in the context of the UNFCCC climate policy instead of, or alongside with, other instruments aimed at reducing deforestation. However, the quality of the data should be improved and some methodological issues, such as the allocation procedure in multiproduct systems and the possible dilution effect through third parties not committed to emission reduction targets, should be considered. - Highlights: Black-Right-Pointing-Pointer CO{sub 2} emissions from land use changes are highly important. Black-Right-Pointing-Pointer Attribution of land use changes for products is difficult. Black-Right-Pointing-Pointer Simple and robust method is developed to attribute land use change emissions.

54 ENVIRONMENTAL SCIENCES↗

Solution of resource allocation problem for identification of cost-effective measures to reduce nuclear proliferation risks

This report presents a methodology of selection of cost-effective measures to reduce nuclear proliferation risks. The methodology relies on a graded security model used in practice in different applications. The method is based on the controlled finite Markov chain approach set in combination with discrete dynamic programming and MCDM (Multi Criteria Decision Making) techniques that enables the expert to select the cost-effective measures to reduce nuclear proliferation risks depending on availability of resources. The analysis performed with different number of possible measures confirms the conclusions that the implementation of extra-large costs may not produce the required effect, and the increase in resources above a certain level does not appear sensitive. Diversification in improving the effectiveness of other measures seems more rational and efficient for the whole system than the unlimited improvement of the effectiveness of only one measure.

98 NUCLEAR DISARMAMENT, SAFEGUARDS, AND PHYSICAL P↗

Inter-well connectivity detection in CO 2 WAG projects using statistical recurrent unit models

Routine well-wise injection and production measurements contain significant information on subsurface structure and properties. Data-driven technology that interprets surface data into subsurface structure or properties can assist operators in making informed decisions by providing a better understanding of field assets. Our machine-learning framework is built on the statistical recurrent unit (SRU) model and interprets well-based injection/production data into inter-well connectivity without relying on a geologic model. We test it on synthetic and field-scale CO 2 EOR projects utilizing the water-alternating-gas (WAG) process. SRU is a special type of recurrent neural network (RNN) that allows for better characterization of temporal trends, by learning various statistics of the input at different time scales. In our application, the complete states (injection rate, pressure and cumulative injection) at injectors and pressure states at producers are fed to SRU as the input and the phase rates at producers are treated as the output. Once the SRU is trained and validated, it is then used to assess the connectivity of each injector to any producer using permutation variable importance method, wherein inputs corresponding to an injector are shuffled and the increase in prediction error at a given producer is recorded as the importance (connectivity metric) of the injector to the producer. This method is tested in both synthetic and field-scale cases. The validation of the proposed data-driven inter-well connectivity assessment is performed using synthetic data from simulation models where inter-well connectivity can be easily measured using the streamline-based flux allocation. The SRU model is shown to offer excellent prediction performance on the synthetic case. Despite significant measurement noise and frequent well shut-ins imposed in the field-scale case, the SRU model offers good prediction accuracy, the overall relative error of the phase production rates at most producers ranges from 10% to 30%. It is shown that the dominant connections identified by the data-driven method and streamline method are in close agreement. This significantly improves confidence in our data-driven procedure. The novelty of this work is that it is purely data-driven method and can directly interpret routine surface measurements to intuitive subsurface knowledge. Furthermore, the streamline-based validation procedure provides physics-based backing to the results obtained from data analytics. This study results in a reliable and efficient data analytics framework that is well-suited for large field applications.

42 ENGINEERING↗

Probabilistic Zonal Reserve Requirements for Improved Energy Deliverability with Wind Power

In power systems with high penetration of renewable energy resources, uncertainty and variability of these stochastic resources introduce additional challenges for the operation of power systems. To improve the power system's reliability in the face of uncertainty, reserves are required as additional generation capacity to rebalance the power system following random disturbances. However, reserve deliverability is not guaranteed, because it may encounter potential transmission line congestion. Zonal reserve requirements can address this issue, but operators lack efficient ways to allocate reserves to zones while accounting for wind power forecast uncertainty. Herein we propose a methodology for probabilistic zonal reserve requirements to address wind power forecast uncertainties. This method estimates the probability distribution of line flows based on the system generation margin and injection shift factor. This estimate is then used to construct pre-defined and post-zonal reserve requirements. Case studies demonstrate that the proposed method efficiently schedules energy and reserves to balance energy and manage deliverability with wind power forecast uncertainty.We also discuss operational implications of the proposed method.

17 WIND ENERGY↗

Better Than Worst-Case Decoding for Quantum Error Correction

The overheads of classical decoding for quantum error correction in cryogenic quantum systems grow rapidly with the number of logical qubits and their correction code distance. Decoding at room temperature is bottlenecked by refrigerator I/O bandwidth while cryogenic on-chip decoding is limited by area/power/thermal budget. To overcome these overheads, we are motivated by the observation that in the common case (over 90% of the time), error correction 'syndromes' are fairly trivial with high redundancy / sparsity, since the error correction codes are over-provisioned to be able to correct for uncommon worst-case complex scenarios (to ensure substantially low logical error rates). If suitably exploited, these trivial scenarios can be handled with insignificant overhead, thereby alleviating any bottlenecks towards handling the worst-case scenarios by state-of-the-art means. Here, we propose Better Than Worst-Case Decoding for Quantum Error Correction, targeting cryogenic quantum systems and Surface Code, consisting of: On-chip Clique Decoder: An extremely lightweight decoder for correcting trivial common-case errors, designed for the cryogenic domain. The decoder is implemented and evaluated for SFQ logic. Statistical Off-chip Bandwidth Allocation: A statistical confidence-based technique for allocation of off-chip decoding bandwidth, to efficiently handle the rare complex decodes that are not covered by the Clique Decoder. Decode-Overflow Execution Stalling: A method to stall circuit execution, for the worst-case scenarios in which the provisioned off-chip bandwidth is insufficient to complete all requested off-chip decodes. In all, BTWC decoding achieves 70-99+% off-chip bandwidth elimination across a range of logical and physical error rates, without significantly sacrificing the accuracy of a state-of-the-art off-chip decoder. Further, it achieves 10-1000x bandwidth reduction over prior bandwidth reduction techniques, as well as 15-37x resource overhead reduction compared to prior on-chip decoding.

cryogenic optimization↗

The Variable Infiltration Capacity model version 5 (VIC-5): infrastructure improvements for new applications and reproducibility

The Variable Infiltration Capacity (VIC) model is a macroscale semi-distributed hydrologic model. VIC development began in the early 1990s and the model has since been used extensively for basin- to global-scale applications that include hydrologic dataset construction, trend analysis of hydrologic fluxes and states, data evaluation and assimilation, forecasting, coupled climate modeling, and climate change impact assessment. Ongoing operational applications of the VIC model include the University of Washington's drought monitoring and forecasting systems and NASA's Land Data Assimilation System. This paper documents the development of VIC version 5 (VIC-5), which includes a major reconfiguration of the legacy VIC source code to support a wider range of modern hydrologic modeling applications. The VIC source code has been moved to a public GitHub repository to encourage participation by the broader user and developer communities. The reconfiguration has separated the core physics of the model from the driver source code, whereby the latter is responsible for memory allocation, preprocessing and post-processing, and input–output (I–O). VIC-5 includes four drivers that use the same core physics modules, but which allow for different methods for accessing this core to enable different model applications. Finally, VIC-5 is distributed with robust test infrastructure, components of which routinely run during development using cloud-hosted continuous integration. The work described here provides an example to the model development community for extending the life of a legacy model that is being used extensively. The development and release of VIC-5 represents a significant step forward for the VIC user community in terms of support for existing and new model applications, reproducibility, and scientific robustness.

42 ENGINEERING↗

Model-Free Building Temperature Control and Power Allocation Under Measurement Time Delays

Taking a step towards a greener planet has created an increased need for a higher integration of renewable energy resources into the electric grid. Nonetheless, the intermittency and uncertainty associated with renewable generation have slowed down this integration. Demand response (DR) has been recently adopted to address this challenge by utilizing demand side flexibility and enabling the participation of many grid-interactive efficient buildings (GEBs). However, existing DR methods require significant modeling and/or training efforts and are computationally expensive. To address the aforementioned issues, we propose a model-free control (MFC)-based strategy that is robust to the time delays in the temperature measurements of the thermostatically controlled loads (TCLs). It assigns to each GEB a local controller to maintain the TCLs’ temperatures within desired comfort levels, while the load aggregator (LA) allocates the assigned reference power provided by the distribution system operator (DSO) to support a specific grid service, such as demand peak reduction, load shifting, balancing supply and demand, and consuming the solar photovoltaic power locally. We investigate the effects of such loss of information on the local control action as well as on meeting the power allocation constraint. We conclude that, for an appropriate choice of design parameters, the proposed MFC controller is satisfactorily robust to measurement time delays.

Telsang, Bhagyashri↗

Adaptive PV Frequency Control Strategy Based on Real-time Inertia Estimation

The declining cost of solar Photovoltaics (PV) generation is driving its worldwide deployment. As conventional generation with large rotating masses is being replaced by renewable energy such as PV, the power system’s inertia will be affected. As a result, the system’s frequency may vary more dramatically in the case of a disturbance, and the frequency nadir may be low enough to trigger protection relays such as under-frequency load shedding. The existing frequency-watt function mandated in power inverters cannot provide grid frequency support in a loss-of-generation event, as PV plants usually do not have power reserves. Here, a novel adaptive PV frequency control strategy is proposed to reserve the minimum power required for grid frequency support. A machine learning model is trained to predict system frequency response under varying system conditions, and an adaptive allocation of PV headroom reserves is made based on the machine learning model as well as real-time system conditions including inertia. Case studies show the proposed control method meets the frequency nadir requirements using minimal power reserves compared to a fixed headroom control approach.

14 SOLAR ENERGY↗

Impact Mapping for Geospatial Reasoning and Decision Making

Objective The reported study evaluated a novel approach to aiding geospatial reasoning and decision making. Background Impact mapping aims to alleviate the cognitive demands of geospatial tasks in part by externalizing data in the form of an integrated decision surface. This is achieved by aggregating data across multiple sources of information and visualizing their combined utility rather than objective measurements or individual utility. Previous research has shown that geospatial decisions improve when aided in this manner, but it remains unknown if dynamic decision making, often plagued by fatigue and anchoring bias, would benefit similarly. Method The experiment implemented a systematic manipulation of the presence of a composite impact map and the number of attributes present in a two-stage disaster relief, resource allocation task to investigate when and how impact mapping is beneficial or deleterious to decision makers. Results The presence of the composite impact map increased the utility of selected sites, increased re-planning decisions, reduced information display views, and reduced workload. Generally, the effect of the composite impact map was greater when participants were asked to evaluate more attributes. Conclusion Composite impact maps appear to improve repeated geospatial reasoning and minimize anchoring bias because they alleviate the cognitive demands otherwise necessary to interpret and maintain information from multiple attributes. Application Data visualization techniques, such as impact mapping, can improve repeated geospatial decision making in environments that include high cognitive demand.

Illingworth, David A.↗

Optoacoustic Microscopy for Investigation of Material Nanostructures-Embracing the Ultrasmall, Ultrafast, and the Invisible

The goal of this grant was the development of a new type of scanning acoustic microscope for nanometer resolution ultrasound imaging, based on ultrafast optoacoustics (>GHz). In the microscope, subpicosecond laser pulses was used to generate and detect very high frequency ultrasound with nanometer wavelengths. We report here on the outcome of the 3-year DOE/BES grant which involved the design, multifaceted construction, and proof-of-concept demonstration of an instrument that can be used for quantitative imaging of nanoscale material features – including features that may be buried so as to be inaccessible to conventional lightwave or electron microscopies. The research program has produced a prototype scanning optoacoustic microscope which, in combination with advanced computational modeling, is a system-level new technology (two patents issues) which offer novel means for precision metrology of material nanostructures, particularly those that are of contemporary interest to the frontline micro- and optoelectronics device industry. For accomplishing the ambitious technical goals, the research roadmap was designed and implemented in two phases. In Phase I, we constructed a “non-focusing” optoacoustic microscope instrument (“POAM”), with nanometer vertical (z-) resolution, while limited to approximately 10 micrometer scale lateral recolution. The Phase I version of the instrument which was guided by extensive acoustic and optical numerical modeling of the basic underlying acoustic and optical physics, featured nanometer scale close loop positioning between the optoacoustic transducer element and a nanostructured material sample under investigation. In phase II, we implemented and demonstrated a scanning version of the instrument (“SOAM”) where incident acoustic energy is focused, and scanned on lateral (x-y) spatial scale in the 100 nm range as per the goals of the project. In so doing we developed advanced numerical simulations to provide computational models of the focusing of multi-GHz acoustic waves to the nanometer scale and innovated a series fabrication approaches for a new type of broadband high-frequency acoustic focusing microscope objective by applying methods on nanoimprinting and focused-ion beam techniques. In the following, the Phase I and Phase II instrument development is reported as Section II. The first segment of this section describes the POAM instrument and its development, while including much of the underlying ultrafast acoustic physics which is common to all of our work for this grant. Then, the science and engineering of the SOAM instrument is described, including the methods of fabricating new types of acoustic microlenses. The results section is followed by reports on publications (Section III), Participants (Section IV), and statement of full use of the allocated grant funds (Section V).

36 MATERIALS SCIENCE↗

A Hierarchical Charging Control of Plug-in Electric Vehicles with Simple Flexibility Model

The demand for plug-in electric vehicles (PEVs) is increasing exponentially, which could have significant effects on the power grid, because charging one PEV is nearly equivalent to adding three houses to the power system. Even though PEVs increase demand on the electric grid, their capability to actively interact with the grid presents excellent opportunities for load management, demand response, and various other grid services. This paper proposes an innovative scheduling and control framework to enable smart PEV charging for grid services while meeting PEV owners’ travel needs. In the proposed framework, each set of electric vehicle supply equipment (EVSE) is equipped with a controller that estimates charging power and energy flexibility based on vehicle characteristics, EVSE power rating, battery energy state, and upcoming trip information. With the charging flexibility model received from each EVSE controller, the central coordinator determines the optimal power allocation for a look-ahead time window for the given grid services. The proposed charging coordination strategy can help reduce the computational complexity and communication requirement compared with existing methods. It is also scalable to the expanding PEV fleet and robust to uncertainties in upcoming vehicle trips and future system condition. The proposed method is studied using a prototypical feeder developed at Pacific Northwest National Laboratory and detailed trip information extracted from the National Household Travel Survey.

Wu, Di↗

Improving User Notification on Frequently Changing HPC Environments

Today s HPC centers user environments can be very complex. Centers often contain multiple large complicated computational systems each with their own user environment. Changes to a system s environment can be very impactful; however, a center s user environment is, in one-way or another, frequently changing. Because of this, it is vital for centers to notify users of change. For users, untracked changes can be costly, resulting in unnecessary debug time as well as wasting valuable compute allocations and research time. Communicating frequent change to diverse user communities is a common and ongoing task for HPC centers. This paper will cover the OLCF s current processes and methods used to communicate change to users of the center s large Cray systems and supporting resources. The paper will share lessons learned and goals as well as practices, tools, and methods used to continually improve and reach members of the OLCF user community.

communication↗

A Novel Coarsening Method for Scalable and Efficient Mesh Generation

In this paper, we propose a novel mesh coarsening method called brick coarsening method. The proposed method can be used in conjunction with any graph partitioners and scales to very large meshes. This method reduces problem space by decomposing the original mesh into fixed-size blocks of nodes called bricks, layered in a similar way to conventional brick laying, and then assigning each node of the original mesh to appropriate brick. Our experiments indicate that the proposed method scales to very large meshes while allowing simple RCB partitioner to produce higher-quality partitions with significantly less edge cuts. Our results further indicate that the proposed brick-coarsening method allows more complicated partitioners like PT-Scotch to scale to very large problem size while still maintaining good partitioning performance with relatively good edge-cut metric. Graph partitioning is an important problem that has many scientific and engineering applications in such areas as VLSI design, scientific computing, and resource management. Given a graph G = (V,E), where V is the set of vertices and E is the set of edges, (k-way) graph partitioning problem is to partition the vertices of the graph (V) into k disjoint groups such that each group contains roughly equal number of vertices and the number of edges connecting vertices in different groups is minimized. Graph partitioning plays a key role in large scientific computing, especially in mesh-based computations, as it is used as a tool to minimize the volume of communication and to ensure well-balanced load across computing nodes. The impact of graph partitioning on the reduction of communication can be easily seen, for example, in different iterative methods to solve a sparse system of linear equation. Here, a graph partitioning technique is applied to the matrix, which is basically a graph in which each edge is a non-zero entry in the matrix, to allocate groups of vertices to processors in such a way that many of matrix-vector multiplication can be performed locally on each processor and hence to minimize communication. Furthermore, a good graph partitioning scheme ensures the equal amount of computation performed on each processor. Graph partitioning is a well known NP-complete problem, and thus the most commonly used graph partitioning algorithms employ some forms of heuristics. These algorithms vary in terms of their complexity, partition generation time, and the quality of partitions, and they tend to trade off these factors. A significant challenge we are currently facing at the Lawrence Livermore National Laboratory is how to partition very large meshes on massive-size distributed memory machines like IBM BlueGene/P, where scalability becomes a big issue. For example, we have found that the ParMetis, a very popular graph partitioning tool, can only scale to 16K processors. An ideal graph partitioning method on such an environment should be fast and scale to very large meshes, while producing high quality partitions. This is an extremely challenging task, as to scale to that level, the partitioning algorithm should be simple and be able to produce partitions that minimize inter-processor communications and balance the load imposed on the processors. Our goals in this work are two-fold: (1) To develop a new scalable graph partitioning method with good load balancing and communication reduction capability. (2) To study the performance of the proposed partitioning method on very large parallel machines using actual data sets and compare the performance to that of existing methods. The proposed method achieves the desired scalability by reducing the mesh size. For this, it coarsens an input mesh into a smaller size mesh by coalescing the vertices and edges of the original mesh into a set of mega-vertices and mega-edges. A new coarsening method called brick algorithm is developed in this research. In the brick algorithm, the zones in a given mesh are first grouped into fixed size blocks called bricks. These brick are then laid in a way similar to conventional brick laying technique, which reduces the number of neighboring blocks each block needs to communicate. Contributions of this research are as follows: (1) We have developed a novel method that scales to a really large problem size while producing high quality mesh partitions; (2) We measured the performance and scalability of the proposed method on a machine of massive size using a set of actual large complex data sets, where we have scaled to a mesh with 110 million zones using our method. To the best of our knowledge, this is the largest complex mesh that a partitioning method is successfully applied to; and (3) We have shown that proposed method can reduce the number of edge cuts by as much as 65%.

97 MATHEMATICS AND COMPUTING↗