Search NASA⌕ Search

Engineering topics

Rose, Amy

Publications and source records attributed to Rose, Amy.

42 records · Page 3

Cross-Border Energy Trade between Nepal and India: Assessment of Trading Opportunities

This study builds on several prior analyses of cross-border energy trading in South Asia to explore the potential trading opportunities between India, Nepal, and the larger South Asia region, including Bangladesh and Bhutan. It uses a detailed one-year, 1-hour resolution unit commitment and economic dispatch model of the South Asia power system to examine the technical and economic impacts of imports and exports under different assumptions about hydropower development in Nepal and regional trade agreements in the year 2022.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗

Cross-Border Electricity Trade between India and Sri Lanka: Impact on Power System Operations

Recent developments in South Asia have the potential to increase the incentives for cross-border energy trade (CBET) between India and Sri Lanka. India's surplus power, including from its rapid expansion of wind and solar generation, provides an incentive for it to increase investment returns through exports. Sri Lanka is considering options to increase its power supply, and CBET could alleviate the need for investments in new thermal generation capacity, while giving it access to India's large and diverse power system. Finally, CBET can facilitate renewable energy (RE) integration by allowing the two countries to more efficiently harness balancing resources and smooth out wind and solar variability over larger geographic areas. This study builds on several prior analyses of a potential high-voltage direct current transmission link between India and Sri Lanka. Uniquely, the study makes use of a detailed one-year, 30-minute resolution unit commitment and economic dispatch (UC&ED) model of both countries to examine the technical and economic impacts of CBET between India and Sri Lanka in the year 2025. The study finds that CBET, enabled by a 500-MW high voltage direct current transmission link, could generate annual production cost savings of USD 180 million and improve power system operations. Through better utilization of the most efficient generators, generation from relatively expensive natural gas plants in Sri Lanka is displaced by lower cost imports from India. In addition, increased flexibility through exports reduces RE curtailment in India's Southern region by 400 GWh annually. While the study captures many previously unconsidered intertemporal operational constraints, it does not provide a complete characterization of system operations with a high-voltage direct current interconnection. Additional work should include a dynamic analysis of system stability following interruptions to the direct current interconnection, an AC power flow analysis and consideration of any non-technical barriers to trade which might prevent the optimization of CBET between India and Sri Lanka at 30-minute timescales.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗

Ramifications of Evolving Misbehaving Convolutional Neural Network Kernel and Batch Sizes

Deep-learners have many hyper-parameters including learning rate, batch size, kernel size - all playing a significant role toward estimating high quality models. Discovering useful hyper-parameter guidelines is an active area of research, though the state of the art generally uses a brute force, uniform grid approach or random search for finding ideal settings. We share the preliminary results of using an alternative approach to deep learner hyper-parameter tuning that uses an evolutionary algorithm to improve the accuracy of a deep-learner models used in satellite imagery building footprint detection. We found that the kernel and batch size hyper-parameters surprisingly differed from sizes arrived at via a brute force uniform grid approach. These differences suggest a novel role for evolutionary algorithms in determining the number of convolution layers, as well as smaller batch sizes in improving deep-learner models.

Coletti, Mark↗

Dynamically-Spaced Geo-Grid Segmentation for Weighted Point Sampling on a Polygon Map Layer

Geo-grid algorithms divide a large polygon area into several smaller polygons, which are important for studying or executing a set of operations on underlying topological features of a map. The current geo-grid algorithms divide a large polygon in to a set of smaller but equal size polygons only (e.g. is ArcMaps Fishnet). The time to create a geo-grid is typically proportional to number of smaller polygons created. This raises two problems - (i) They cannot skip unwanted areas (such as water bodies, given about 71% percent of the Earth's surface is water-covered); (ii) They are incognizant to any underlying feature set that requires more deliberation. In this work, we propose a novel dynamically spaced geo-grid segmentation algorithm that overcomes these challenges and provides a computationally optimal output for borderline cases of an uneven polygon. Our method uses an underlying topological feature of population distributions, from the LandScan Global 2016 dataset, for creating grids as a function of these weighted features. We benchmark our results against available algorithms and found our approach improves geo-grid creation. Later on, we demonstrate the proposed approach is more effective in harvesting Points of Interest data from a crowd-sourced platform.

Sims, Kelly↗

Building Extraction at Scale Using Convolutional Neural Network: Mapping of the United States

Establishing up-to-date large scale building maps is essential to understand the urban dynamics, such as estimating population, urban planning, and many other applications. Although many computer vision tasks have been successfully carried out with deep convolutional neural networks, there is a growing need to understand their large scale impact on building mapping with remote sensing imagery. In this paper, taking advantage of the scalability of convolutional neural networks (CNNs) and using only few areas with the abundance of building footprints, for the first time we conduct a comparative analysis of four state-of-the-art CNNs for extracting building footprints across the entire continental United States. The four CNN architectures namely: Branch-out CNN, fully convolutional network (FCN), conditional random field as recurrent neural network (CRFasRNN), and SegNet, support semantic pixelwise labeling and focus on capturing textural information at multiscale. We use 1-meter resolution aerial images from National Agriculture Imagery Program as the test-bed, and compare the extraction results across the four methods. In addition, we propose to combine signed-distance labels with SegNet, the preferred CNN architecture identified by our extensive evaluations, to advance building extraction results to instance level. We further demonstrate the usefulness of fusing additional near IR information into the building extraction framework. Large scale experimental evaluations are conducted and reported using metrics that include: Precision, recall rate, intersection over union, and the number of buildings extracted. With the improved CNN model and no requirement of further postprocessing, we have generated building maps for the United States with an average processing time less than one minute for an area of size &#x223C;<!-- ~ --> 56 km 2 . Finally, the quality of extracted buildings and processing time demonstrated that the proposed CNN based framework fits the need of building extraction at scale.

97 MATHEMATICS AND COMPUTING↗

Spatializing Global Urban Extent: A Source Driven Approach

“Urban” is something that intuitively feels very well defined; however, when it comes time to express this idea on a map, things get complicated. Definitions of “urban” vary globally, and as such there is not universal understanding of what makes a given place “urban”. The common approach to spatially defining urban extents is through remotely sensed imagery. The alternative approach presented in this paper uses the percent of population in urban areas, which is a common macroeconomic (country-level) variable with the definition for urban generally defined by each country’s statistical office, along with temporally-aligned highresolution population data to spatially define urban extents for each country. Because the percent urban number is defined by the same producer as other urban/rural defined statistical data, such as household characteristics or birth rate, understanding the spatial aspect of urban from the same perspective is ideal for high resolution spatial modeling of these other phenomena.

Moehl, Jessica↗