Navegando por Assunto "Parameter estimation"

Agora exibindo 1 - 3 de 3

Acesso aberto (Open Access)
Modelagem chuva-vazão-produção de sedimentos via problemas inversos
(Universidade Federal do Pará, 2023-10-05) TORRES FALCÓN, Cindy; ESTUMANO, Diego Cardoso; http://lattes.cnpq.br/5521162828533153; https://orcid.org/0000-0003-4318-4455; BLANCO, Claudio José Cavalcante; http://lattes.cnpq.br/8319326553139808; https://orcid.org/0000-0001-8022-2647
The development of mathematical models and direct methods has made it possible to predict hydrological phenomena such as rainfall-runoff-sediment yield. In order to complement the simulation model, inverse problems can be used to determine the properties of these phenomena and estimate parameters that cannot be measured directly. Therefore, this study was carried out in a small catchment in the Amazon with precipitation data and parameters estimated using the Kineros2 (K2) / direct model (DM). The study proposes solutions to the inverse problem (IP) characterized by the rainfall-runoff-sediment yield phenomenon for events with scarce data, with the aim of estimating the inflow rate, estimating the physical parameters, the runoff depth and the sediment yield of the basin analyzed. The sediment yield data comes from the sediment gauge station in the small catchment. For a more precise and detailed analysis of the model's behavior, combinations of information from observations and the K2 model were also carried out simultaneously with IP. The main scientific contribution is the application of the inverse problem method (Bayesian inference together with a Fourier series) to estimate the parameters of the kinematic wave model and the mass balance, and to estimate the runoff depth and sediment yield for a small watershed in the Amazon. The results showed a good fit between the observed and predicted data via IP, as Nash-Sutcliffe coefficients above 0.70 and RMSE between 0.27 and 1.99 were obtained in the calibration and validation of the rainfall-runoff-sediment yield model. The simulation of the runoff depth and sediment yield showed a 95% degree of reliability, which is consistent with the observed data.
Acesso aberto (Open Access)
Modelagem, simulação e estimação dos parâmetros por MCMC de um modelo que descreve a dinâmica de adsorção em uma coluna de leito fixo
(Universidade Federal do Pará, 2023-01-18) SOEIRO, Wilhamis Fonseca; VIEGAS, Bruno Marques; http://lattes.cnpq.br/1196600058247902; HTTPS://ORCID.ORG/0000-0002-2768-652X; ESTUMANO, Diego Cardoso; http://lattes.cnpq.br/5521162828533153; https://orcid.org/0000-0003-4318-4455
The treatment of industrial effluents is extremely important for both the environment and human health. The purification of water from polluting components, such as metals and organic compounds, for reuse in the industrial process can be considered one of the main applications in this field. Therefore, there is interest in modeling one of the most used treatment processes, adsorption. Aiming to describe the dynamics of the process in an adsorption column, in this work the method of lines and the pdepe function (matlab) are used to solve the model formed by the mass balance in the liquid phase, linear driving force equation (LDF) and the Langmuir isotherm for equilibrium. An evaluation of the model varying some experiment conditions was carried out, from which results congruent with those found in the literature were observed. In addition, a sensitivity analysis of the phenomenon was carried out in relation to the parameters: Langmuir constant, intraparticle mass transfer coefficient and axial dispersion coefficient. Subsequently, these parameters were estimated using the Monte Carlo technique via Markov chain (MCMC) using experimental data found in the literature. Finally, in general, the estimates were good enough to represent the adsorption dynamics of the evaluated experiments.
Acesso aberto (Open Access)
Redes neurais de múltiplas camadas para redução do tempo de aquisição de dados para testes modais em estruturas flexíveis
(Universidade Federal do Pará, 2007-05-14) MACHADO, José Aristides dos Santos; MELLO, Hiran de; http://lattes.cnpq.br/2127559774805521; VIEIRA JÚNIOR, Petrônio; http://lattes.cnpq.br/1958791286192330
This work presents techniques to improve the dynamics parameters estimation of Single Degree of Freedom (SDOF) and Multiple Degree of Freedom (MDOF) models characteristic from flexible structures. The analyses are referred to the method that uses the Frequency Response Function (FRF) obtained from the impulse response of the flexible structure. We use for assumption that the considered models are convenient for a suitable description of the system. Thus, an experimental good method of obtaining the FRF should produce a significant accordance between the theoretical and the experimental FRF. The improvement in increasing the acquisition time artificially (forecasting) is analyzed by using a Multilayer Neural Network (MNN) model. The performance of neural forecaster is compared with results obtained using ARX and ARMAX models. The obtained results in this research, suggest the viability to use the MNN.