چكيده به لاتين
In recent years, the use of electric vehicles due to environmental issues and the high efficiency of engines electric has been increasing in comparison with combustion engines. One of the important and key components in electric vehicles is the battery management system. Among these storage sources, lithium-ion batteries have attracted a lot of attention in portable devices, especially electric vehicles, due to their high energy density, long life, suitable nominal voltage, low self-discharge and reasonable cost. The battery management system not only measures the battery charge level accurately, but also ensures the safe operation of the vehicle and increases the battery life. Accurately estimating the charge level of a lithium-ion battery is a very challenging task because the lithium-ion battery is an electrochemical system that is highly non-linear. Since it is not possible to calculate the battery charge level directly, various estimation methods have been developed for this purpose. Among them, the use of model-based methods are particularly popular due to their simplicity and less expensive calculations and high accuracy. In this thesis, Tonen's exact model with two branches (RC) is used. Which has high accuracy despite relatively low calculations. And to determine the parameters, their variable value at different temperature and charge level is used, and the uncertainty caused by the modeling error is included in it. To reduce and eliminate high frequency noises, we use wavelet transform and to estimate the charge level, we use sliding controller, sliding controller with adaptive gain, and adaptive sliding controller methods. With this method, the removal of high-frequency noises with the least time and the least cost of the first stage and reducing the impact of disturbances and uncertainty of the model and parameters in the second stage is done by the controllers. High accuracy, robustness and simplicity of implementation are other features of this method. And to evaluate their performance and validation, the data published by McMaster University in Canada was used.
