Bibliographic citations
Vera, E., Vargas, G. (2023). Evaluación hidrológica con uso de redes neuronales artificiales para el cálculo de hidrogramas en la cuenca alta del Huallaga [Trabajo de Suficiencia Profesional, Universidad Peruana de Ciencias Aplicadas (UPC)]. http://hdl.handle.net/10757/671144
Vera, E., Vargas, G. Evaluación hidrológica con uso de redes neuronales artificiales para el cálculo de hidrogramas en la cuenca alta del Huallaga [Trabajo de Suficiencia Profesional]. PE: Universidad Peruana de Ciencias Aplicadas (UPC); 2023. http://hdl.handle.net/10757/671144
@misc{renati/411779,
title = "Evaluación hidrológica con uso de redes neuronales artificiales para el cálculo de hidrogramas en la cuenca alta del Huallaga",
author = "Vargas Valerio , Giancarlo Anthony",
publisher = "Universidad Peruana de Ciencias Aplicadas (UPC)",
year = "2023"
}
The project has been executed in the Huallaga basin, which is approximated using a neural network method, given that it has a history of accumulating large amounts of precipitation and flows, which generate outlet hydrographs with considerable flow, leaving the streams where they are in a state of partial collapse, this has generated problems in the population, which is looking for a quick solution, and a flood detection method can be carried out through a conventional hydrological analysis, however the calculation of floods usually takes a long time. time, with respect to the reaction time that one has in the face of an emergency, that is where a neural network method is proposed, which can carry out this procedure quickly, without considering large errors, for the project the Huallaga basin has been modeled a total number of 250 times, generating a training dataset, which consisted of a relationship between the input hyetograph and an output hydrograph, which through a regression could directly approximate the result, given that it had already performed training of the flows generated for low, medium and strong rainfall, the error achieved is 8.17%, having been trained for a total of 2000 iterations, considering an mse error metric, with an Adam optimizer and an MLP type architecture , finally managing to extrapolate the hydrograph values for the outlet of the basin.
This item is licensed under a Creative Commons License