| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 4.34 MB | Adobe PDF |
Authors
Advisor(s)
Abstract(s)
A evolução das tecnologias de armazenamento de energia está, nos dias de hoje, associada à recorrente transformação do atual paradigma da produção da energia elétrica e à forma como é consumida. Por questões de necessidade, e não apenas por conveniência, é cada vez mais inevitável o armazenamento da energia elétrica que não pode ser consumida no momento em que é produzida, para garantir sustentabilidade energética. Portanto, por revelarem alta densidade energética, alta eficiência e uma maior longevidade, as baterias de iões de lítio são uma tecnologia de armazenamento de energia que têm sido recentemente aplicadas no armazenamento de energia proveniente da produção de origem renovável, bem como no sector dos veículos elétricos/híbridos e nos dispositivos portáteis.
Assim sendo, este trabalho tem como objetivo o desenvolvimento de um método de carregamento, que tem por base um algoritmo de inteligência artificial designado como redes neuronais, que privilegie a minimização da deterioração precoce das baterias com o propósito de prolongar o seu tempo de vida útil. Para a elaboração deste método proposto o trabalho desenvolvido pode ser descrito em duas fases.
A primeira é referente à análise e estudo da resistência interna de baterias com diferentes níveis de envelhecimento, onde serão comparados e verificados os valores da resistência interna em várias condições de carregamento. Isto possibilitará uma concepção real do estado de vida das baterias e também do modo como estas se comportam durante um carregamento.
A segunda fase consiste no desenvolvimento de um método de carregamento com o recurso à implementação de uma rede neuronal. O método proposto minimizará a deterioração das baterias e garantirá um carregamento adaptado e ideal para as suas condições. Os resultados obtidos são comparados com métodos tradicionais existentes na literatura.
The evolution of energy storage systems is nowadays associated with the recurrent transformation of the current paradigm of electric energy production and the way it is consumed. For reasons of necessity, and not just for convenience, it is increasingly unavoidable to store electrical energy that cannot be consumed when it is produced to guarantee energy sustainability. Therefore, as they demonstrate high energy density, high efficiency and longer longevity, lithium-ion batteries are an energy storage technology that recently has been applied to the storage of energy from renewable sources, as well as in the vehicle electric / hybrid sector and portable devices. Therefore, this work has the objective of developing a charging method, based on an artificial intelligence algorithm called neural networks, which focuses on the minimization of early deterioration of the batteries with the purpose of prolonging their useful life. For the elaboration of this proposed method the work developed can be reconciled in two phases. The first one refers to the analysis and study of the internal resistance of batteries with different levels of aging, where the values of the internal resistance for various conditions will be compared and verified. This will allow a real conception of the state of life of the batteries as well as the way it behaves during a charging. The second phase consists in the development of a charging method with the implementation of a neural network. The proposed method will minimize the deterioration of batteries and guarantee an adapted and ideal charge for their conditions. The results obtained are compared with traditional methods in the literature.
The evolution of energy storage systems is nowadays associated with the recurrent transformation of the current paradigm of electric energy production and the way it is consumed. For reasons of necessity, and not just for convenience, it is increasingly unavoidable to store electrical energy that cannot be consumed when it is produced to guarantee energy sustainability. Therefore, as they demonstrate high energy density, high efficiency and longer longevity, lithium-ion batteries are an energy storage technology that recently has been applied to the storage of energy from renewable sources, as well as in the vehicle electric / hybrid sector and portable devices. Therefore, this work has the objective of developing a charging method, based on an artificial intelligence algorithm called neural networks, which focuses on the minimization of early deterioration of the batteries with the purpose of prolonging their useful life. For the elaboration of this proposed method the work developed can be reconciled in two phases. The first one refers to the analysis and study of the internal resistance of batteries with different levels of aging, where the values of the internal resistance for various conditions will be compared and verified. This will allow a real conception of the state of life of the batteries as well as the way it behaves during a charging. The second phase consists in the development of a charging method with the implementation of a neural network. The proposed method will minimize the deterioration of batteries and guarantee an adapted and ideal charge for their conditions. The results obtained are compared with traditional methods in the literature.
Description
Keywords
Baterias Li-Ion Redes Neuronais. Resistência Interna Tecnologias de Armazenamento de Energia Elétrica