Eloy Kaviski

Monthly rainfall–runoff modelling using artificial neural networks

Abstract Rainfall–runoff models usually present good results, but parameter calibration sometimes is tedious and subjective, and in many cases it depends on additional data surveys in the field. An alternative to the conceptual models is provided by empirical models, which relate input and output by means of an arbitrary mathematical function that bears no direct relationship to the physical characteristics of the rainfall–runoff process. This category includes the artificial neural networks (ANNs), whose implementation is the main focus of this paper. This study evaluated the capacity of ANNs to model with accuracy the monthly rainfall–runoff process. The case study was performed in the Jangada River basin, Paraná, Brazil. The results of the three ANNs that produced the best results were compared to those of a conceptual model at monthly time scale, IPHMEN. The ANNs presented the best results with highest correlation coefficients and Nash-Sutcliffe statistics and the smallest difference of volume. Citation Machado, F., Mine, M., Kaviski, E. & Fill, H. (2011) Monthly rainfall–runoff modelling using artificial neural networks. Hydrol. Sci. J. 56(3), 349–361.

Aplicação do método AHP para a hierarquização das alternativas de tratamento de resíduos sólidos domiciliares

The current situation of the Metropolitan Region of Curitiba (RMC) in terms of the management of domestic solid waste was the initial motivation for this study. It aims to identify and evaluate alternative technologies for municipal solid waste treatment in order to minimize the quantity of solid waste disposed in landfills. The results resulted from the analysis of the amount of domestic solid waste produced in RMC, and also on a literature review, which was used to identify some technologies used to treat municipal solid waste. The AHP - Analytic Hierarchy Process was used to establish a coherent hierarchical arrangement of the alternative treatments. The application of this method placed recycling at the top of the hierarchy, with a value for the sum of weights of 3,488, whereas the value for incineration was 1,005. Those results show that it is possible to determine, in a decreasing order, the position of each technology to be used for the treatment of municipal solid waste in addition to sanitary landfills. The main result of the study was the production of tool that allowed the simulation of sceneries based on the alteration of the values of the criteria adopted. Each scenario allowed the utilization of the same method applied to RMC to support decision making in other circumstances, which can be specific for each municipality or a consortium of municipalities.

Modelagem do fluxo de pedestres pela teoria macroscópica

The need to model the traffic of pedestrians has become a priority in environments where pedestrians can circulate. The research has advanced in an attempt to solve accidents related to large crowds. Furthermore, analyzing structures for pedestrians before they have been built has become a requirement of the authorities, as well as being able to manage the flow of pedestrians in those facilities. To characterize this behavior, studies have emerged including theories of physics and mathematics, which analyze pedestrian traffic through models described by differential and integral-differential equations. This work presents the macroscopic scale modeling of traffic and its direct connection with physics and hydrodynamic models. The model was applied to a walk situation where the results were compared with other work on the simulations and experimental observations on the pedestrian traffic, found in the bibliography.

Curvas de permanência de qualidade da água como subsídio para o enquadramento de corpos d'água a partir de modelagem matemática em regime não permanente

Modelos matematicos de qualidade de agua tem sido utilizados como ferramentas para o planejamento e a gestao de recursos hidricos em diversos paises. Em geral, em rios, as analises sao baseadas em simulacoes considerando condicoes de escoamento de regime permanente. Entretanto, bacias urbanas apresentam uma intensa dinâmica, consequente de mecanismos de uso e ocupacao do solo, capaz de induzir processos de poluicao que requerem uma analise que contemple aspectos fisicos, quimicos e biologicos. A abordagem em estado nao-permanente permite descrever de forma mais realista estas caracteristicas, e associar de maneira consistente as interacoes de qualidade e quantidade de agua, em funcao da representacao hidrodinâmica de um sistema. Esta pesquisa contribui para esse entendimento usando como estudo de caso o rio Iguacu, localizado na regiao metropolitana de Curitiba, ao analisar potenciais impactos de avaliacoes de qualidade baseadas em uma comparacao de condicoes de estado permanente e nao-permanente como forma de subsidiar elementos para a gestao do uso da agua. O modelo aqui proposto (SIHQUAL), desenvolvido em modulos tradicionais, utiliza metodos de diferencas finitas explicitas para solucao das equacoes de Saint-Venant integradas com a equacao de transporte de massa, para demanda bioquimica de oxigenio (DBO) e oxigenio dissolvido (OD). Para a analise de qualidade da agua utilizou-se do conceito de curva de permanencia de qualidade da agua. Observou-se principalmente que modelos em estado transiente oferecem maior flexibilidade e uma distinta perspectiva para o gerenciamento, em especial quanto a aplicacao dos instrumentos de gestao de recursos hidricos, como o enquadramento.

Simulação de escoamento de água em canal entre placas paralelas utilizando o método do reticulado de Boltzmann

O metodo do reticulado de Boltzmann (LBM) possui equacao governante caracterizada pelas etapas de transmissao e colisao, e pode ser visto como uma forma discretizada da equacao cinetica de Boltzmann em que somente sao mantidos os detalhes moleculares essenciais para recuperar o comportamento macroscopico. O metodo e eficiente na simulacao de escoamento de fluidos, mostrando-se competitivo nas aplicacoes de mecanica dos fluidos computacional, sendo especialmente util em aplicacoes com geometrias complexas. O objetivo deste trabalho e utilizar o LBM para simular o escoamento de agua em um canal entre placas paralelas. Para isto, foram usadas condicoes de contorno bounce back e Zou-He. Fez-se a comparacao entre as solucoes numerica e analitica, para a validacao do metodo.

Modelagem computacional do tráfego de veículos pela teoria microscópica

The necessity to model the flow of vehicles has become a priority in large urban centers and transit routes in general. This type of modeling can be used as a tool to establish new strategies, policy management and traffic control, to improve efficiency in the transport system. The vehicle traffic can be physically represented by three scales: microscopic, kinetic and macroscospic. In mathematical modeling of complex systems, such as traffic, none of these scale representations is effectively consistent with the physics of the system analyzed. On the other hand, cellular automata are distinct among the main computational models for simulation of vehicular traffic due to its ability to represent the principal characteristics of the traffic by using simple rules and low computational cost. The difference is that this type of modeling does not involve any kind of equation. The main objective of this work is to understand the use of computational modeling in the study of vehicular traffic through cellular automata. This is done by presenting a model that employs microscopic rules on movement of vehicles on the track with a single one-way road. The results are expressed through by graphs, flow vs. density, known as the fundamental diagram, and density vs. velocity.

Modelagem e Escoamento do Tráfego de Véıculos em uma Pista Simples

O aumento significativo no fluxo do trafego de veiculos tem causado diversos problemas para a sociedade. O estudo das caracteristicas do trafego de veiculos tem se tornado uma ferramenta importante para avaliar e planejar as rodovias. Nesse contexto, a simulacao computacional via automatos celulares pode ser utilizada como uma importante ferramenta para obtencao de caracteristicas fundamentais do trafego em rodovias ou em areas urbanas. Neste trabalho, desenvolveu-se um modelo de automatos celulares que utilizam regras simples de movimentacao dos veiculos, para analisar o tempo de escoamento de uma pista simples. A partir dos dados obtidos, formulou-se um modelo de regressao polinomial que relaciona o tempo medio total de escoamento e o numero de veiculos. Os resultados apresentados mostraram-se bastante satisfatorios e podem ser usados para desenvolver ferramentas de controle ou informacao sobre o trafego.

DAILY SCHEDULING OF SMALL HYDRO POWER PLANTS DISPATCH WITH MODIFIED PARTICLES SWARM OPTIMIZATION

This paper presents a new approach for short-term hydro power scheduling of reservoirs using an algorithm-based Particle Swarm Optimization (PSO). PSO is a population-based algorithm designed to find good solutions to optimization problems, its characteristics have encouraged its adoption to tackle a variety of problems in different fields. In this paper the authors consider an optimization problem related to a daily scheduling of small hydro power dispatch. The goal is construct a feasible solution that maximize the cascade electricity production, following the environmental constraints and water balance. The paper proposes an improved Particle Swarm Optimization (PSO) algorithm, which takes advantage of simplicity and facility of implementation. The algorithm was successfully applied to the optimization of the daily schedule strategies of small hydro power plants, considering maximum water utilization and all constraints related to simultaneous water uses. Extensive computational tests and comparisons with other heuristics methods showed the effectiveness of the proposed approach.