Determinação de parâmetros ótimos da equação intensidade, duração e frequência de chuvas utilizando otimização heurística

Imagem de Miniatura



Título da Revista

ISSN da Revista

Título de Volume


Universidade Federal de Goiás


Considering that the intensity, duration and frequency curves (IDF curves) are scarce and in some cases outdated for the State of Goiás, the present work aims to elaborate an algorithm capable of treating the rainfall data, coming from INMET rain gauges, and, later, through the use of heuristic algorithms, to create a program capable of determining the parameters $K$, $a$, $b$ and $c$ through the rainfall disaggregation method. The probabilistic distribution used was the Gumbel distribution, which was evaluated through the Kolmogorov-Smirnov adherence test. Then, the algorithms: genetic, ant colony, particle swarm, flower pollination and differential evolution were used to determine the parameters. Subsequently, an adjustment test was performed using the adjusted determination coefficient, Wilmott's concordance index, and the confidence coefficient to measure the performance of the proposed models. The results obtained showed that the algorithms are able to solve the problem effectively, with emphasis on the flower pollination algorithm and the differential evolution algorithm, which presented a low standard deviation and high processing speed. Regarding the adjusted determination coefficient, Willmott's concordance index and the confidence coefficient, none of them presented values lower than 0.9945, 0.9950 and 0.9923, respectively. Thus, the model and the parameters obtained effectively describe the rain data for the analyzed cities, receiving an optimal performance rating. Finally, it is concluded that the tool elaborated in this work fulfills the proposed objective, serving not only as a practical calculation tool, but also as study material.



IDF, Chuva, Algoritmo heurístico, Modelagem, IDF, Rain, Heuristic algorithm, Modeling


SANTOS NETO, E. P. Determinação de parâmetros ótimos da equação intensidade, duração e frequência de chuvas utilizando otimização heurística. 2020. 109 f. Dissertação (Mestrado em Modelagem e Otimização) - Universidade Federal de Goiás, Catalão, 2020.