Estimação do estado de carga de baterias de lítio-íon em diferentes condições de temperaturas utilizando filtros de Kalman

• Autor(a) principal: Santos, Rafael de Melo Silva
• Data de Publicação: 2019
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Biblioteca Digital de Teses e Dissertações da UFPB
• Texto Completo: https://repositorio.ufpb.br/jspui/handle/123456789/19499
• topic: Bateria de lítio; Processamento digital de sinais; Filtros de Kalman; Modelagem de baterias; Sistema de gerenciamento de baterias; Estimação de SoC de baterias; Lithium-ion battery; Digital signal processing; Kalman filters; Battery modeling; BMS; Battery SoC estimation; CNPQ::ENGENHARIAS::ENGENHARIA ELETRICA

One of the challenges of humanity in the twenty-first century is to reverse environmental degradation in progress, without halting human development. This has been called, very appropriately, sustainable development. To achieve sustainable development, technological innovation in both the generation and efficient use of energy is imperative. The development of electric vehicles is a key part of this sustainable development, and consequently the use of batteries, given that they are the elements responsible for storing electrical energy in electrochemical form. Estimating the State of Charge (SoC) is fundamental for the dissemination of this type of transportation, among other current applications. In this context, a methodology was developed that allows the estimation of charge of batteries for different temperature conditions using Kalman filters, especially the extended Kalman filter, which allows to evaluate the behavior of SoC for a range of 10 C, 25 C and 40 C. In addition, a method for identifying the equivalent circuit model of a battery is proposed. Based on the methodology developed, some preliminary results were obtained with the use of a Stationary Lithium-ion (LiFePO4) battery of 20 Ah capacity and nominal voltage of 3.3 V. From the proposed method to identify the appropriate equivalent circuit model of the battery and the parameters of this model, it was possible to determine a model with three RC branches for the Lithium-ion battery and to estimate the parameters of this model. Then the state of charge for temperatures of 10 C, 25 C and 40 C were determined. From numerical results, the extended Kalman Filter used in the research showed convergence in the prediction of the SoC, in which the covariance error is between 3 % and 6 %, showing the the proposed method convergence. Otherwise, the Kalman filter reached a maximum residual of 83 %.