Navegando por Autor "GOMES, Evanice Pinheiro"

Agora exibindo 1 - 2 de 2

Acesso aberto (Open Access)
Avaliação de modelos de inteligência artificial híbridos na estimativa de precipitações
(Universidade Federal do Pará, 2022-03-18) GOMES, Evanice Pinheiro; BLANCO, Claudio José Cavalcante; http://lattes.cnpq.br/8319326553139808
The hydrological analyzes carried out from rainfall in the Amazon are essential due to its importance in climate regulation, regional and global atmospheric circulation. However, in this region, there are limitations related to data series with short periods and many flaws, especially in the daily scale. Despite significant advances in science and technology, practical and accurate predictions have been a major concern due to their complexity. Therefore, several conceptual models, empirical or hybrid, have been tested to forecast rain with greater precision. Among empirical models, those that incorporate artificial intelligence (AI) methods are potentially useful approaches to simulate the precipitation process. Artificial Neural Networks (ANN), as AI models, are able to establish a relationship between historical inputs (rain, flow, etc.) and the desired outputs, through a non-linear function composed of several factors that are adjusted to the observed data, allowing your prediction. Thus, to improve the precipitation analysis, hybrid models were developed, involving Artificial Neural Network (ANN) of the type with Time Delay (TDNN), ELMAN network, Radial Base network (RBF) and Adaptive Neuro-Fuzzy Inference System (ANFIS), coupled with Maximum Overlap Discrete Wavelet (MODWT). Six rainfall gauge station were adopted, which are located in different biomes of the region, and satellite data (CMORPH). Rainfall data were evaluated by seasonal periods (rainy and dry). The results obtained demonstrated that the MODWT-ANFIS model had the best capacity to simulate the daily precipitation of the evaluated rainfall gauge station, even for dry periods, which are known to be more difficult to be simulated in relation to the rainy periods. In this case, data entries lagged by 4 days and 5 days performed better, with Nash values close to 1.0 and root mean square errors below 0.001.
Acesso aberto (Open Access)
Regionalização de precipitações via fuzzy C-means
(Universidade Federal do Pará, 2017-04-05) GOMES, Evanice Pinheiro; BLANCO, Claudio José Cavalcante; http://lattes.cnpq.br/8319326553139808
The knowledge of the precipitation behavior is indispensable, since any change in its quantity and spatial and temporal distributions have an important impact on nature and consequently on the various human activities. However, in precipitation studies, the lack of rainfall monitoring, generating the lack of information over time and spatially in the river basins is a problem for the understanding of this variable. In order to overcome this problem, the rainfall regionalization method was proposed. The main idea was to divide the Tocantins Araguaia - RHTA hydrographic region into homogeneous regions, defined by the Fuzzy C-means method. The Euclidean distance was adopted as a measure of similarity and the fuzzification parameter, ranging from 1.2 to 2.0, and the explanatory variables of rainfall (altitude, latitude and longitude) were used as input data. Three homogeneous regions were obtained, which were validated by the PBM index and the heterogeneity test. The frequencies of observed rainfall events were generated for the 83 rain gauge stations, distributed in their respective regions, and calibrated by the Normal, Log-Normal, Gama, Gumbel, Exponential, Logarithmic and Weibull probability functions. With the application of the chi-square test, we defined the best probability function in the occurrence of rainfall in each homogeneous region. The validation of the probabilities functions was performed in 9 target stations, using the chi-square test. In this stage, it was observed that for annual average precipitation, data adherence occurred to all the rain gauge stations, since they presented results of the chi-square test of less than 5.99 (for Log-normal distribution functions). It was also observed that for monthly average precipitation, data were adhered to all the rainfall stations with the Gama and Weibull functions. For the simulation of rain depth, Linear, Potential, Exponential and Logarithm models were tested through the multiple regression method, using as independent variables, altitude, latitude and longitude. As performance criterion of the models, the R², R²_a, E%, ε%, NASH and RMSE were used. In the simulation of annual averages, the Linear model presented the best performance indices. In the estimation ofviii monthly averages, all multiple regression models did not perform well, with errors above 50%, which motivated the estimation of monthly rainfall for rainy and dry periods. In this new approach the regression models presented excellent performance criteria with errors below 10%. The performance indexes allowed us to conclude that the regional models developed for the homogeneous regions of rainfall, defined by the Fuzzy C-Means method, are a good option in the estimation of annual and monthly average rainfall and are important for a better understanding of the rainfall regime in RHTA, and can serve as a tool for better planning of water resources in the region.