College of Engineering

Permanent URI for this community

https://hdl.handle.net/1969.6/93658

Browse

Now showing 1 - 20 of 210

Assessing the effects of irrigation water salinity on two ornamental crops by remote spectral imaging
(MDPI, 2/20/2021) Yu, Xinyang; Her, Younggu; Chang, Anjin; Song, Jung-Hun; Campoverde, E. Vanessa; Schaffer, Bruce
Salinity is one of the most common and critical environmental factors that limit plant growth and reduce crop yield. The aquifers, the primary sources of irrigation water, of south Florida are shallow and highly permeable, which makes agriculture vulnerable to projected sea level rise and saltwater intrusion. This study evaluated the growth responses of two ornamental nursery crops to the different salinity levels of irrigation water to help develop saltwater intrusion mitigation plans for the improved sustainability of the horticultural industry in south Florida. Two nursery crops, Hibiscus rosa-sinensis and Mandevilla splendens, were treated with irrigation water that had seven different salinity levels from 0.5 (control) to 10.0 dS/m in the experiment. Crop height was measured weekly, and growth was monitored daily using the normalized difference vegetation index (NDVI) values derived from multispectral images collected using affordable sensors. The results show that the growth of H. rosa-sinensis and M. splendens was significantly inhibited when the salinity concentrations of irrigation water increased to 7.0 and 4.0 dS/m, for each crop, respectively. No significant differences were found between the NDVI values and plant growth variables of both H. rosa-sinensis and M. splendens treated with the different irrigation water salinity levels less than 2.0 dS/m. This study identified the salinity levels that could reduce the growth of the two nursery crops and demonstrated that the current level of irrigation water salinity (0.5 dS/m) would not have significant adverse effects on the growth of these crops in south Florida.
COVID-19 vaccine hesitancy: The role of socioeconomic factors and spatial effects
(MDPI, 2/24/2022) Lee, Jim; Huang, Yuxia
This paper investigates the spatial dimension of socioeconomic and demographic factors behind COVID-19 vaccine hesitancy. With a focus on a county with considerable sociodemographic diversity in the state of Texas, USA, we apply regression models to census-tract-level data of the unvaccinated population. In addition to disparities in accessing the vaccination service, particularly for residents in rural areas, empirical results confirm under-vaccination among lower socioeconomic neighborhoods and communities with signs of distrust in government. The spatial model regressions further underscore the impact that vaccine hesitancy among residents in one community spread to its nearby communities. This observed spatial spillover effect is attributable to the geographic interactions of similar socioeconomic groups.
Massive sensor array fault tolerance: Tolerance mechanism and fault injection for validation
(Hindawi, 2010-08-24) Um, Dugan
As today's machines become increasingly complex in order to handle intricate tasks, the number of sensors must increase for intelligent operations. Given the large number of sensors, detecting, isolating, and then tolerating faulty sensors is especially important. In this paper, we propose fault tolerance architecture suitable for a massive sensor array often found in highly advanced systems such as autonomous robots. One example is the sensitive skin, a type of massive sensor array. The objective of the sensitive skin is autonomous guidance of machines in unknown environments, requiring elongated operations in a remote site. The entirety of such a system needs to be able to work remotely without human attendance for an extended period of time. To that end, we propose a fault-tolerant architecture whereby component and analytical redundancies are integrated cohesively for effective failure tolerance of a massive array type sensor or sensor system. In addition, we discuss the evaluation results of the proposed tolerance scheme by means of fault injection and validation analysis as a measure of system reliability and performance.
Space-aware data integration for ocean observing systems
(Elsevier, 2012-01-08) Li, Longzhuang; Nalluri, Anil kumar; Ai, Lirong
Spatial features are important properties with respect to data integration in many areas such as ocean observational information and environmental decision making. In order to address the needs of these applications, we have to represent and reason about the spatial relevance of various data sources. In this paper, we develop and implement a space-aware data integration prototype system to facilitate retrieval and integration of data from the in situ ocean observing stations in the Gulf of Mexico. The prototype system adopt the state-of-the-art qualitative spatial representation and reasoning techniques to represent partonomic, distance, and topological, and directional relations to answer spatial-related queries.
Spray-on PEDOT:PSS and P3HT:PCBM thin films for polymer solar cells
(MDPI, 2014-01-21) Eslamian, Morteza; Newton, Joshua E.
PEDOT:PSS electron-blocking layer, and PEDOT:PSS + P3HT:PCBM stacked layers are fabricated by ultrasonic atomization and characterized by scanning electron microscopy (SEM) and optical profilometry. The measured thicknesses based on SEM and optical profilometry are quite different, indicating the incapability of measurement techniques for non-uniform thin films. The thickness measurements are compared against theoretical estimations and a qualitative agreement is observed. Results indicate that using a multiple pass fabrication strategy results in a more uniform thin film. It was also found that the film characteristics are a strong function of solution concentration and spraying passes, and a weak function of substrate speed. Film thickness increases with solution concentration but despite the prediction of theory, the increase is not linear, indicating a change in the film porosity and density, which can affect physical and opto-electrical properties. Overall, while spray coating is a viable fabrication process for a wide range of solar cells, film characteristics can be easily altered by a change in process parameters.
GIS-based tools to calculate class location and maximum allowable operating pressure (MAOP) for gas transmission pipelines
(2015-05) Lide, Anna G.
Two GIS-based analysis tools were developed for gas transmission pipelines in this project. One tool determines class location designations along a gas transmission pipeline to satisfy the U.S. Department of Transportation (DOT) regulations. The second tool calculates the maximum allowable operating pressure (MAOP) based on the structural properties of pipeline segments. The DOT regulations in 49 CFR 192 explain the class location definitions and the operating pressure as a function of class location design factors. The DOT class locations indicate a level of the potential risk of damage to a pipeline from external activities and the potential risk to people and property near the pipeline. Class location categories set the parameters that affect the MAOP values for operating a gas pipeline. The categories are a function of population and infrastructure in proximity to the pipeline. Specifying the MAOP value reduces the risks to the pipeline and surrounding area.
2D and 3D Mapping of a Littoral Zone with UAS and Structure from Motion Photogrammetry
(2015-05) Giessel, Justin Zachary
Advancements in the miniaturization of sensors and their integration in light‐weight, smallscale unmanned aerial systems (UAS) have resulted in an explosion of uses for inexpensive and easily obtained remotely sensed data. This study examines the capabilities of a small‐scale UAS equipped with a consumer grade RGB camera for 2D and 3D mapping of a sandy bay shoreline using Structure from Motion (SfM) photogrammetry. Several key components are analyzed in order to assess the utility of UAS‐based SfM photogrammetry for beach and boundary surveying of the littoral zone. First, the accuracy of the 3D point cloud produced by the SfM densification process over the beach is compared to high accuracy RTK GPS transects. Results show a mean agreement of approximately 7.9 cm over the sub‐aerial beach with increased error in shallow water. Minimal effects of beach slope on vertical accuracy were observed. Secondly, bathymetric measurements extracted from the UAS/SfM point cloud are examined, and an optical inversion approach is implemented where the SfM method fails. Results show that a hybrid elevation model of the beach and littoral zone consisting of automatic SfM products, post‐processed SfM products, and optical inversion provide the most accurate results when mapping over turbid water. Finally, SfM‐derived shoreline elevation contour (boundary) is compared to a shoreline elevation contour derived using the currently accepted RTK GPS method for conducting legal littoral boundary surveys in the state of Texas. Results show mean planimetric offsets < 25 cm demonstrating the potential of UAS‐based SfM photogrammetry for conducting littoral boundary surveys along non‐occluded, sandy shorelines.
New achievements in control of robotic systems
(Hindawi, 2015-05-28) Santibáñez, Víctor; Zavala-Río, Arturo; Garcia Carrillo, Luis Rodolfo; Moreno-Valenzuela, Javier
Nowadays, robotics is experiencing a noticeable growth mainly due to the successful partnership between theory and practice, whose alliance distinguishes the modern robotics from early robotics. This progress is the result of the interplay between the engineering and scientific communities of different disciplines. In this sense, control engineering plays a major role not only in the development of new robotic systems, but also in the performance improvement of existing and traditional systems. The robot control can be considered as the cornerstone of robotics. Its task consists in driving a robot to accomplish a desired task, in a fully autonomous way. Advanced methods of control have been required in order to adequately face challenging problems which have arisen from underactuation, visual servoing, visual tracking, locomotion, cooperative manipulation, among others. The aim of this special issue has been to collect novel and original works from the automatic control community in the area of robotic systems, such as industrial robots, legged robots, robot manipulators, cooperative manipulators, and multiagent systems. We hope this collection of papers will provide important and original information to roboticists and control systems researchers. We also consider that the publication of these theoretical, numerical, and experimental contributions will lead to solving some of the challenges that the development of new robotic system demands.
Reciprocal estimation of pedestrian location and motion state toward a smartphone geo-context computing solution
(MDPI, 2015-06-15) Liu, Jingbin; Zhu, Lingli; Wang, Yunsheng; Liang, Xinlian; Hyyppä, Juha; Chu, Tianxing; Liu, Keqiang; Chen, Ruizhi
The rapid advance in mobile communications has made information and services ubiquitously accessible. Location and context information have become essential for the effectiveness of services in the era of mobility. This paper proposes the concept of geo-context that is defined as an integral synthesis of geographical location, human motion state and mobility context. A geo-context computing solution consists of a positioning engine, a motion state recognition engine, and a context inference component. In the geo-context concept, the human motion states and mobility context are associated with the geographical location where they occur. A hybrid geo-context computing solution is implemented that runs on a smartphone, and it utilizes measurements of multiple sensors and signals of opportunity that are available within a smartphone. Pedestrian location and motion states are estimated jointly under the framework of hidden Markov models, and they are used in a reciprocal manner to improve their estimation performance of one another. It is demonstrated that pedestrian location estimation has better accuracy when its motion state is known, and in turn, the performance of motion state recognition can be improved with increasing reliability when the location is given. The geo-context inference is implemented simply with the expert system principle, and more sophisticated approaches will be developed.
Nano sensing and energy conversion using surface plasmon resonance (SPR)
(MDPI, 2015-07-16) Kim, Iltai (Isaac); Kihm, Kenneth David
Nanophotonic technique has been attracting much attention in applications of nano-bio-chemical sensing and energy conversion of solar energy harvesting and enhanced energy transfer. One approach for nano-bio-chemical sensing is surface plasmon resonance (SPR) imaging, which can detect the material properties, such as density, ion concentration, temperature, and effective refractive index in high sensitivity, label-free, and real-time under ambient conditions. Recent study shows that SPR can successfully detect the concentration variation of nanofluids during evaporation-induced self-assembly process. Spoof surface plasmon resonance based on multilayer metallo-dielectric hyperbolic metamaterials demonstrate SPR dispersion control, which can be combined with SPR imaging, to characterize high refractive index materials because of its exotic optical properties. Furthermore, nano-biophotonics could enable innovative energy conversion such as the increase of absorption and emission efficiency and the perfect absorption. Localized SPR using metal nanoparticles show highly enhanced absorption in solar energy harvesting. Three-dimensional hyperbolic metamaterial cavity nanostructure shows enhanced spontaneous emission. Recently ultrathin film perfect absorber is demonstrated with the film thickness is as low as ~1/50th of the operating wavelength using epsilon-near-zero (ENZ) phenomena at the wavelength close to SPR. It is expected to provide a breakthrough in sensing and energy conversion applications using the exotic optical properties based on the nanophotonic technique.
Inferring human activity in mobile devices by computing multiple contexts
(MDPI, 2015-08-28) Chen, Ruizhi; Chu, Tianxing; Liu, Keqiang; Liu, Jingbin; Chen, Yuwei
This paper introduces a framework for inferring human activities in mobile devices by computing spatial contexts, temporal contexts, spatiotemporal contexts, and user contexts. A spatial context is a significant location that is defined as a geofence, which can be a node associated with a circle, or a polygon; a temporal context contains time-related information that can be e.g., a local time tag, a time difference between geographical locations, or a timespan; a spatiotemporal context is defined as a dwelling length at a particular spatial context; and a user context includes user-related information that can be the user’s mobility contexts, environmental contexts, psychological contexts or social contexts. Using the measurements of the built-in sensors and radio signals in mobile devices, we can snapshot a contextual tuple for every second including aforementioned contexts. Giving a contextual tuple, the framework evaluates the posteriori probability of each candidate activity in real-time using a Naïve Bayes classifier. A large dataset containing 710,436 contextual tuples has been recorded for one week from an experiment carried out at Texas A&M University Corpus Christi with three participants. The test results demonstrate that the multi-context solution significantly outperforms the spatial-context-only solution. A classification accuracy of 61.7% is achieved for the spatial-context-only solution, while 88.8% is achieved for the multi-context solution.
Adaptive neural network-based event-triggered control of single-input single-output nonlinear discrete-time systems
(IEEE, 2015-10-26) Sahoo, Avimanyu; Xu, Hao; Sarangapani, Jagannathan
This paper presents a novel adaptive neural network (NN) control of single-input and single-output uncertain nonlinear discrete-time systems under event sampled NN inputs. In this control scheme, the feedback signals are transmitted, and the NN weights are tuned in an aperiodic manner at the event sampled instants. After reviewing the NN approximation property with event sampled inputs, an adaptive state estimator (SE), consisting of linearly parameterized NNs, is utilized to approximate the unknown system dynamics in an event sampled context. The SE is viewed as a model and its approximated dynamics and the state vector, during any two events, are utilized for the event-triggered controller design. An adaptive event-trigger condition is derived by using both the estimated NN weights and a dead-zone operator to determine the event sampling instants. This condition both facilitates the NN approximation and reduces the transmission of feedback signals. The ultimate boundedness of both the NN weight estimation error and the system state vector is demonstrated through the Lyapunov approach. As expected, during an initial online learning phase, events are observed more frequently. Over time with the convergence of the NN weights, the inter-event times increase, thereby lowering the number of triggered events. These claims are illustrated through the simulation results.
Comparison of linear and non-linear feature extraction on vegetation and oil spill hyperspectral images
(2015-12) Ramirez-Aguilar, Andres
A hyperspectral image provides a multidimensional figure rich in data consisting of hundreds of spectral dimensions. For this research, the method of analysis for a hyperspectral image will consist of two different feature extraction algorithms: principal component analysis locally linear embedding. Analyzing the spectral and spatial information of such image with linear and non-linear algorithms will result in high computational time. In order to overcome this problem, this research proposes a system using a MapReduce-Graphics Processing Unit (GPU) model that can help analyze a hyperspectral image through the usage of parallel hardware and a parallel programming model, which will be simpler to handle compared to other low level parallel programming models. Additionally, Hadoop will be used as an open-source version of the MapReduce parallel programming model. The ultimate goal of this research is to provide a foundation for a simple and powerful system that is scalable and easily extend-able.
Experimental verification of epsilon-near-zero plasmon polariton modes in degenerately doped semiconductor nanolayers
(Optica Publishing Group, 2016) Campione, Salvatore; Kim, Iltai; de Ceglia, Domenico; Keeler, Gordon A.; Luk, Ting S.
We investigate optical polariton modes supported by subwavelength-thick degenerately doped semiconductor nanolayers (e.g. indium tin oxide) on glass in the epsilon-near-zero (ENZ) regime. The dispersions of the radiative (R, on the left of the light line) and non-radiative (NR, on the right of the light line) ENZ polariton modes are experimentally measured and theoretically analyzed through the transfer matrix method and the complex-frequency/real-wavenumber analysis, which are in remarkable agreement. We observe directional near-perfect absorption using the Kretschmann geometry for incidence conditions close to the NR-ENZ polariton mode dispersion. Along with field enhancement, this provides us with an unexplored pathway to enhance nonlinear optical processes and to open up directions for ultrafast, tunable thermal emission.
Approximate optimal control of affine nonlinear continuous-time systems using event-sampled neurodynamic programming
(IEEE, 2016-04-07) Sahoo, Avimanyu; Xu, Hao; Sarangapani, Jagannathan
This paper presents an approximate optimal control of nonlinear continuous-time systems in affine form by using the adaptive dynamic programming (ADP) with event-sampled state and input vectors. The knowledge of the system dynamics is relaxed by using a neural network (NN) identifier with event-sampled inputs. The value function, which becomes an approximate solution to the Hamilton–Jacobi–Bellman equation, is generated by using event-sampled NN approximator. Subsequently, the NN identifier and the approximated value function are utilized to obtain the optimal control policy. Both the identifier and value function approximator weights are tuned only at the event-sampled instants leading to an aperiodic update scheme. A novel adaptive event sampling condition is designed to determine the sampling instants, such that the approximation accuracy and the stability are maintained. A positive lower bound on the minimum inter-sample time is guaranteed to avoid accumulation point, and the dependence of inter-sample time upon the NN weight estimates is analyzed. A local ultimate boundedness of the resulting nonlinear impulsive dynamical closed-loop system is shown. Finally, a numerical example is utilized to evaluate the performance of the near-optimal design. The net result is the design of an event-sampled ADP-based controller for nonlinear continuous-time systems.
Designing geodatabases for the general authority for statistics of the Kingdom of Saudi Arabia
(2016-05) Alghamdi, Khalid Abdullah; Huang, Yuxia
Developing a new system of both statistical surveys and geographic entities for the General Authority for Statistics (GaStat) in Saudi Arabia is needed to respond to fast growing statistical data as well as to be more relevant with users and related resources through providing a diversity of datasets. This project developed a new geodatabase conceptual model for the GaStat by adopting the methods used by the United States Census Bureau. The new model consists of two main components: statistical surveys and geographic entities. First, statistical surveys use the methods of field surveys, partnership agreements, and self-response to feed the database in the Information Bank, which is a data warehouse and will be used in GaStat. More specifically, additional types of statistical surveys are identified and included in the proposed model. These types include the Saudi Community Survey, Saudi Housing Survey, Saudi Income Survey, Saudi Spending Survey, Saudi Economic Survey, Saudi Industry Survey, Saudi Agricultural Survey, and the Saudi Employment Survey. Further, a series of formatting time methods including quarters, one year, three years, and five years is used in the statistical systems in order to provide up-to-date information. Second, similar to the geographic entities used in U.S. Census Bureau, geographic entities for the GaStat are classified into two groups: legal and administrative entities, and statistical entities based on their corresponding geographic subdivision, and they are further organized in a hierarchical structure. This design should provide a powerful tool for collecting information and creating a standard set for the GaStat in order to retrieve the requests from users. More specifically, the legal and administrative entities include the country and its provinces, governorates, holy areas, economic zones, ZIP code areas, school districts and voting districts. Statistical entities include regions, statistical tracts, statistical block groups, statistical blocks, urban areas, urban growth areas, places, sample-data areas and governorates subdivisions. Additionally, a unique reference numeric code is used to integrate between the data from both statistical surveys and geographic entities. The proposed geodatabase model is expected to address the limitations of the current systems in the GaStat.
Comparison of airborne surveying techniques for mapping submerged objects in shallow water
(2016-08) Nazeri, Behrokh; Starek, Michael J.; Smith, Richard; Jeffress, Gary A.
In this study, bathymetric lidar, high resolution aerial imagery, and hyperspatial resolution imagery collected from a small unmanned aircraft system (UAS) were examined in order to delineate submerged objects in shallow coastal water. A region surrounding Shamrock Island in Corpus Christi Bay along the Texas Gulf Coast was chosen for this study. This area is significant because of the existence of submerged structures including oil pipelines, which may influence the marine environment and navigation in shallow water. Therefore, mapping submerged structures is the first step of any further study in this area in terms of environmental litter and navigation hazards. Different methods were compared to each other in these categories in terms of efficiency and accuracy to map the bathymetric surface and detect submerged structures. First, three different interpolation methods including 2D Delaunay triangulated irregular network (TIN), inverse distance weight (IDW), and multilevel B spline were used to create digital elevation models (DEMs) using airborne lidar data to investigate their use on submerged pipeline detection. Then three different algorithms including Sobel, Prewitt, and Canny were examined in edge detection image processing to illustrate the potential pipelines using aerial imagery. To improve visibility, glint correction methods were implemented and compared to non-glint corrected imagery for pipeline delineation. Finally, a small UAS equipped with a digital camera was flown to evaluate structure from motion (SfM) photogrammetry for bathymetric mapping in the shallow bay. Methods examined included glint corrected imagery and single bands vs. original multiband imagery. The goal was to determine the effectiveness of image pre-conditioning methods for improving UAS-SfM mapping of submerged bottom and structures in shallow water. Results showed that B-spline interpolation method was the best fit compared to other methods for deriving bathymetric DEMs from the airborne lidar data. In edge detection image processing, Canny method performed better between all three methods in detecting the pipelines in the aerial imagery. In the last part, using glint removal methods and green single band imagery as inputs into the UAS-SfM photogrammetry workflow increased the quality of the produced point cloud over shallow water in terms of point density and depth estimation respectively. In conclusion, bathymetric lidar data in fusion with aerial imagery improved the pipeline delineation. Due to inherent limitations in current bathymetric lidar system resolvance power, it is recommended that future surveys targeted for this objective plan as best as possible for ideal water conditions in terms of visibility, employ more scan overlap. Sun glint correction improved the quality of the imagery in terms of penetrating through the water column. Avoiding sun glint by choosing appropriate place and time for data collection is the best way to deal with sun glint. In the UAS-SfM part, using a polarized filter on RGB cameras is recommended to assess the sun glint effect in the result.
Cotton growth modeling and assessment using unmanned aircraft system visual-band imagery
(SPIE, 2016-08-23) Chu, Tianxing; Chen, Ruizhi; Landivar, Juan; Maeda, Murilo; Yang, Chenghai; Starek, Michael J.
This paper explores the potential of using unmanned aircraft system (UAS)-based visible-band images to assess cotton growth. By applying the structure-from-motion algorithm, the cotton plant height (ph) and canopy cover (cc) information were retrieved from the point cloud-based digital surface models (DSMs) and orthomosaic images. Both UAS-based ph and cc follow a sigmoid growth pattern as confirmed by ground-based studies. By applying an empirical model that converts the cotton ph to cc, the estimated cc shows strong correlation (R2=0.990) with the observed cc. An attempt for modeling cotton yield was carried out using the ph and cc information obtained on June 26, 2015, the date when sigmoid growth curves for both ph and cc tended to decline in slope. In a cross-validation test, the correlation between the ground-measured yield and the estimated equivalent derived from the ph and/or cc was compared. Generally, combining ph and cc, the performance of the yield estimation is most comparable against the observed yield. On the other hand, the observed yield and cc-based estimation produce the second strongest correlation, regardless of the complexity of the models.
A new temperature-vegetation triangle algorithm with variable edges (TAVE) for satellite-based actual evapotranspiration estimation
(MDPI, 2016-09-07) Zhang, Hua; Gorelick, Steven M.; Avisse, Nicolas; Tilmant, Amaury; Rajsekhar, Deepthi; Yoon, Jim
The estimation of spatially-variable actual evapotranspiration (AET) is a critical challenge to regional water resources management. We propose a new remote sensing method, the Triangle Algorithm with Variable Edges (TAVE), to generate daily AET estimates based on satellite-derived land surface temperature and the vegetation index NDVI. The TAVE captures heterogeneity in AET across elevation zones and permits variability in determining local values of wet and dry end-member classes (known as edges). Compared to traditional triangle methods, TAVE introduces three unique features: (i) the discretization of the domain as overlapping elevation zones; (ii) a variable wet edge that is a function of elevation zone; and (iii) variable values of a combined-effect parameter (that accounts for aerodynamic and surface resistance, vapor pressure gradient, and soil moisture availability) along both wet and dry edges. With these features, TAVE effectively addresses the combined influence of terrain and water stress on semi-arid environment AET estimates. We demonstrate the effectiveness of this method in one of the driest countries in the world—Jordan, and compare it to a traditional triangle method (TA) and a global AET product (MOD16) over different land use types. In irrigated agricultural lands, TAVE matched the results of the single crop coefficient model (−3%), in contrast to substantial overestimation by TA (+234%) and underestimation by MOD16 (−50%). In forested (non-irrigated, water consuming) regions, TA and MOD16 produced AET average deviations 15.5 times and −3.5 times of those based on TAVE. As TAVE has a simple structure and low data requirements, it provides an efficient means to satisfy the increasing need for evapotranspiration estimation in data-scarce semi-arid regions. This study constitutes a much needed step towards the satellite-based quantification of agricultural water consumption in Jordan.
Thunderstorm predictions using artificial neural networks
(2016-10-19) Tissot, Philippe
Artificial neural network (ANN) model classifiers were developed to generate ≤ 15 h predictions of thunderstorms within three 400-km2 domains. The feed-forward, multilayer perceptron and single hidden layer network topology, scaled conjugate gradient learning algorithm, and the sigmoid (linear) transfer function in the hidden (output) layer were used. The optimal number of neurons in the hidden layer was determined iteratively based on training set performance. Three sets of nine ANN models were developed: two sets based on predictors chosen from feature selection (FS) techniques and one set with all 36 predictors. The predictors were based on output from a numerical weather prediction (NWP) model. This study amends an earlier study and involves the increase in available training data by two orders of magnitude. ANN model performance was compared to corresponding performances of operational forecasters and multi-linear regression (MLR) models. Results revealed improvement relative to ANN models from the previous study. Comparative results between the three sets of classifiers, NDFD, and MLR models for this study were mixed—the best performers were a function of prediction hour, domain, and FS technique. Boosting the fraction of total positive target data (lightning strikes) in the training set did not improve generalization.

Browse

Browsing College of Engineering by Issue Date

Results Per Page

Sort Options