Publication:
A weather nowcasting approach as a service for solar energy generation forecast

dc.contributor.advisor Rodríguez Solís, Rafael A.
dc.contributor.author Minotta Zapata, Felipe
dc.contributor.college College of Engineering en_US
dc.contributor.committee León Colón, Leyda V.
dc.contributor.committee Andrade Rengifo, Fabio
dc.contributor.committee Arzuaga, Emmanuel
dc.contributor.department Department of Electrical and Computer Engineering en_US
dc.contributor.representative Ríos, Isabel
dc.date.accessioned 2022-05-24T12:05:28Z
dc.date.available 2022-05-24T12:05:28Z
dc.date.issued 2022-04-05
dc.description.abstract This work focuses on the problem of forecasting the energy availability of solar panel arrangements in smart grids. To solve this problem, we propose to use weather radar information to make a 15 minute ahead prediction of the weather conditions in the area where the renewable sources are located. To this end, we propose to use a framework in which the classical approaches, object- and area-based methods, can work in coop fashion to obtain a reliable forecast system. Each weather system is identified and tracked as an object with multiple features. Meanwhile, the movement direction is obtained by a wind vector field generated by the area-based approach. The framework uses the historical information of each system such as average reflectivity and size change to produce an estimated change in both features, as well as the mean movement direction established by the wind vector field to produce a weather forecast. Python was the programming language selected for the implementation which allows portability and integration with different radar applications. Experiments made with TropiNet weather radar network historical data showed similar metric performance as related nowcasting approaches. en_US
dc.description.abstract Este trabajo se enfoca en el problema de pronosticar la disponibilidad de energía de arreglos de paneles solares en redes inteligentes. Para resolver este problema, proponemos utilizar la información de radares meteorológicos para hacer una predicción de las condiciones climáticas en 15 minutos en el área donde se encuentran las fuentes renovables. Con este fin, proponemos utilizar un entorno en el que los enfoques clásicos, métodos basados en objetos y en área, pueden funcionar de manera cooperativa para obtener un sistema de pronóstico confiable. Cada sistema meteorológico se identifica y rastrea como un objeto con múltiples características. Mientras tanto, la dirección del movimiento se obtiene mediante un campo vectorial de viento generado por el enfoque basado en áreas. El entorno utiliza la información histórica de cada sistema, como la reflectividad promedio y el cambio de tamaño, para producir un cambio estimado en ambas características, así como también la dirección de movimiento promedio establecida por el campo de vectores de viento para producir un pronóstico del tiempo. Python fue el lenguaje de programación seleccionado para la implementación, ya que permite la portabilidad e integración con diferentes aplicaciones de radares. Los experimentos realizados con datos históricos de la red de radares meteorológicos TropiNet mostraron un métricas con rendimiento similar a enfoques relacionados de predicción inmediata. en_US
dc.description.graduationSemester Spring en_US
dc.description.graduationYear 2022 en_US
dc.description.sponsorship Investigation subsidized with funds from the National Science Foundation ACI-1541106 en_US
dc.identifier.uri https://hdl.handle.net/20.500.11801/2894
dc.language.iso en en_US
dc.rights Attribution-NonCommercial-NoDerivs 3.0 United States *
dc.rights.holder (c) 2022 Felipe Minotta Zapata en_US
dc.rights.uri http://creativecommons.org/licenses/by-nc-nd/3.0/us/ *
dc.subject Nowcasting (Meteorology) – Puerto Rico en_US
dc.subject Renewable sources en_US
dc.subject Smart power grids – Puerto Rico en_US
dc.subject.lcsh Probability forecasts (Meteorology) en_US
dc.subject.lcsh Tropical meteorology en_US
dc.subject.lcsh Nowcasting (Meteorology) en_US
dc.subject.lcsh Smart power grids – Puerto Rico en_US
dc.subject.lcsh Meteorological stations, Radar – Puerto Rico en_US
dc.subject.lcsh Weather radar networks – Puerto Rico en_US
dc.subject.lcsh Weather forecasting – Puerto Rico en_US
dc.subject.lcsh Solar panels – Puerto Rico
dc.title A weather nowcasting approach as a service for solar energy generation forecast en_US
dc.type Dissertation en_US
dspace.entity.type Publication
thesis.degree.discipline Electrical Engineering en_US
thesis.degree.level Ph.D. en_US
Files
Original bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
INEL_MinottaZapataF_2022.pdf
Size:
1.52 MB
Format:
Adobe Portable Document Format
Description:
License bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
license.txt
Size:
2.26 KB
Format:
Item-specific license agreed upon to submission
Description: