Pronóstico de la demanda eléctrica residencial basado en el modelo de regresión adaptativa mulltivariante spline (MARS)
A problem to be solved much interest in engineering is to determine the behavior of the electricity demand of users over time, because this does not have a linear characteristic and is also subject to a wide variety of exogenous variables, including the prevailing weather conditions, time of year, d...
| Autor Principal: | Ortiz Alvarado, Miguel Eduardo |
|---|---|
| Formato: | bachelorThesis |
| Idioma: | spa |
| Publicado: |
2016
|
| Materias: | |
| Acceso en línea: |
http://dspace.ups.edu.ec/handle/123456789/11290 |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| Sumario: |
A problem to be solved much interest in engineering is to determine the behavior of the electricity demand of users over time, because this does not have a linear characteristic and is also subject to a wide variety of exogenous variables, including the prevailing weather conditions, time of year, demographic and economic variables, as well as an inherent randomness in individual use. The prediction of energy demand of consumers has the purpose to improve reliability, make decisions and system planning, among others. The most
6
appropriate method to resolve this issue is not clearly defined, however, this work is carried out a general review of the models and methods commonly employed for prognosis, with emphasis on technical Multivariate Adaptive Regression Splines (MARS) which it is a nonparametric regression model and its main objective is to forecast the values of a dependent variable or outcome of a set of independent or predictor variables. Also take note that you must have sufficient historical information on the daily load curve for statistical and forecasting studies, in order to plan future resources and will be a tool for the implementation of an intelligent network because improve the characteristics of demand management and management tasks in distribution. Applying the model research proposes to define a pattern of residential consumption historical data and extrapolate to future predictions. |
|---|