A carregar...
Publicação
Estimativa do caudal em rios através de tratamento de imagem
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 3.35 MB | Adobe PDF |
Autores
Orientador(es)
Resumo(s)
A Velocimetria de Imagens em Grande Escala (LSPIV) baseia-se na análise de uma sequência
de imagens, permitindo medir o campo de velocidades 2D à superfície de um curso de água. Para esse
efeito é necessária a existência de traçadores na superfície da água que sejam solidárias com a
corrente. Isso pode passar pela necessidade de existência de partículas sólidas (detritos de plantas,
pequenos flutuadores – naturais ou adicionados –, bolhas ou padrões de turbulência) de modo a serem
reconhecidos de uma imagem para a outra.
A medição de caudais é essencial para vários aspetos, nomeadamente na gestão das
necessidades de consumo de recursos hídricos, no tratamento de águas residuais, na produção de
energia elétrica, na aferição das taxas de reservas de água nos aquíferos e na prevenção de
inundações. Para que isto seja possível, em Portugal são efetuadas várias medições de caudais nas
linhas de água. Contudo, nem sempre as condições são favoráveis para a realização destas medições.
O método agora proposto oferece uma grande vantagem em relação aos métodos convencionais: não
é intrusivo, ou seja, não irá perturbar o ecossistema em estudo nem alterar o caudal aquando da captura
de vídeo. Outra vantagem deste método em comparação com os métodos convencionais é o facto de
este poder ser usado em situações extremas (como cheias ou perto de estruturas hidráulicas – como
barragens ou pontes) sem apresentarem grandes riscos tanto para o trabalhador como para o material.
Para comprovar este método realizaram-se ensaios em três canais do Pavilhão de Hidráulica
Fluvial e Hidráulica de Estruturas do Departamento de Hidráulica e Ambiente do Laboratório Nacional
de Engenharia Civil e um ensaio de campo num curso de água de maiores dimensões no Carregado,
Portugal.
O estudo consistiu em examinar o tamanho e o formato dos traçadores que forneciam valores
de velocidade superficial mais exatos e precisos. Utilizou-se traçadores quadrangulares de papel com
1, 2, 4, e 8 cm de lado e ainda o poliestireno. Verificou-se que quanto mais pequeno o traçador melhor
é a caracterização do movimento superficial da água, sendo que, o melhor traçador foi o poliestireno.
Como o poliestireno é prejudicial ao ambiente considerou-se que traçadores quadrangulares de papel
com, no máximo, 4 cm de lado são adequados para a aplicação do método LSPIV.
Além do estudo aos traçadores, foi também efetuado um estudo ao coeficiente necessário
aplicar às velocidades superficiais para obter as velocidades médias em altura. Concluiu-se que o
coeficiente varia entre 0.85 e 1 e ainda que o seu valor aumenta com o aumento da altura de água,
diminui com o aumento da rugosidade fundo e diferentes secções de escoamento possuem diferentes
coeficientes.
Para aplicar o método LSPIV é necessário um drone com uma câmara digital incorporada,
traçadores artificiais, caso os naturais não sejam suficientes, e um computador com um software de
análise LSPIV. Para examinar as gravações captadas, nesta dissertação, foram utilizados dois
softwares: o PIVlab e o Fudaa. Outro dos objetivos foi determinar qual destes softwares fornece
resultados mais exatos e precisos. O PIVlab foi aquele que apresentou resultados mais coerentes em
todos os casos de estudo.
Large-scale Image Velocimetry (LSPIV) is based on the analysis of a sequence of images, allowing the measurement of the 2D velocity field at the surface of a watercourse. For this purpose, it is necessary to have contrasts on the water surface that are solidary with the flow. This may involve the need for solid particles (plant debris, small floats - natural or added -, bubbles or turbulence patterns) to be treated from one image to another. Flow measurement is essential, namely in the management of the needs for the consumption of water resources, in the treatment of wastewater, in the production of electric energy, in the measurement of the rates of water reserves in aquifers and in the prevention of floods. To make this possible, in Portugal several flow measurements are made in rivers and waterbeds. However, the conditions of the rivers aren’t always favorable for these measurements. The proposed method offers a great advantage over conventional methods: it isn’t intrusive, i.e., it won’t disturb the ecosystem under study or change the flow rate when capturing a video. Another advantage of this method compared to conventional methods is that it can be used in extreme situations (such as floods or near hydraulic structures - such as dams or bridges) without presenting big risks to the worker and the material. To prove this method, tests were carried out on three channels of the Pavilion of Fluvial Hydraulics and Structural Hydraulics of the Department of Hydraulics and Environment of the National Civil Engineering Laboratory and a field test on a water course in Carregado, Portugal. The study consisted of examining the size and shape of the tracers that provided the most accurate and precise surface velocity values. Square paper tracers with 1, 2, 4, 8 cm sides and polystyrene were used. It was found that the smaller the tracer, the better the characterization of the surface movement of the water. The best tracer was polystyrene. As polystyrene is harmful to the environment, it was considered square paper tracers, with a maximum of 4 cm on each side, are suitable for the application of the LSPIV method. In addition to the study of tracers, was also carried out a study on the coefficient necessary to apply to the surface velocities to obtain the speed average. It was concluded that the coefficient is in between 0.85 and 1, the coefficient value increases with the increase of the water height, decreases with the increase of deep roughness and different flow sections have different coefficients. To apply the LSPIV method, you need a drone with an incorporated digital camera, artificial tracers, if the natural ones are not enough, and a computer with LSPIV analysis software. To examine how recordings were captured, in this dissertation, two software’s were used: PIVlab and Fudaa. Another objective was to determine which of these software offers more accurate and precise results. The PIVlab was the one that presented the most coherent results in all studies cases.
Large-scale Image Velocimetry (LSPIV) is based on the analysis of a sequence of images, allowing the measurement of the 2D velocity field at the surface of a watercourse. For this purpose, it is necessary to have contrasts on the water surface that are solidary with the flow. This may involve the need for solid particles (plant debris, small floats - natural or added -, bubbles or turbulence patterns) to be treated from one image to another. Flow measurement is essential, namely in the management of the needs for the consumption of water resources, in the treatment of wastewater, in the production of electric energy, in the measurement of the rates of water reserves in aquifers and in the prevention of floods. To make this possible, in Portugal several flow measurements are made in rivers and waterbeds. However, the conditions of the rivers aren’t always favorable for these measurements. The proposed method offers a great advantage over conventional methods: it isn’t intrusive, i.e., it won’t disturb the ecosystem under study or change the flow rate when capturing a video. Another advantage of this method compared to conventional methods is that it can be used in extreme situations (such as floods or near hydraulic structures - such as dams or bridges) without presenting big risks to the worker and the material. To prove this method, tests were carried out on three channels of the Pavilion of Fluvial Hydraulics and Structural Hydraulics of the Department of Hydraulics and Environment of the National Civil Engineering Laboratory and a field test on a water course in Carregado, Portugal. The study consisted of examining the size and shape of the tracers that provided the most accurate and precise surface velocity values. Square paper tracers with 1, 2, 4, 8 cm sides and polystyrene were used. It was found that the smaller the tracer, the better the characterization of the surface movement of the water. The best tracer was polystyrene. As polystyrene is harmful to the environment, it was considered square paper tracers, with a maximum of 4 cm on each side, are suitable for the application of the LSPIV method. In addition to the study of tracers, was also carried out a study on the coefficient necessary to apply to the surface velocities to obtain the speed average. It was concluded that the coefficient is in between 0.85 and 1, the coefficient value increases with the increase of the water height, decreases with the increase of deep roughness and different flow sections have different coefficients. To apply the LSPIV method, you need a drone with an incorporated digital camera, artificial tracers, if the natural ones are not enough, and a computer with LSPIV analysis software. To examine how recordings were captured, in this dissertation, two software’s were used: PIVlab and Fudaa. Another objective was to determine which of these software offers more accurate and precise results. The PIVlab was the one that presented the most coherent results in all studies cases.
Descrição
Palavras-chave
LSPIV Velocidade média velocidade superficial coeficiente traçadores e caudal