دانلود پایان نامه دکتری
عنوان :
Intelligent Battery Management System for
Electric Vehicles
زبان : انگلیسی
سال انتشار : 2010
در زمینه : مهندسی برق
کد محصول : 1202
تعداد صفحات : 204
قیمت :25000 تومان
چکیده انگلیسی :
The automotive industry has experienced a significant boom in recent years, accelerating the problems of energy shortage and environmental disruption around the world. To solve the two problems, electric vehicles (EVs), including battery electric vehicles (BEV), hybrid electric vehicles (HEV), and fuel-cell electric vehicles (FEV), have been proposed and studied in recent years. Despite the efforts devoted to the development of EVs by both the scientific research and industrial communities, there are still many obstacles hindering the mass commercialization of EVs. Among these obstacles, the battery system, the new energy storage component in EVs, is one of the most important yet most difficult parts of EV design, and the battery management system (BMS) is recognized as the single most important technical issue in the successful commercialization of EVs.
A vehicular battery must consist of a large number of cells to provide the necessary energy and power. Management only at the level of the battery pack causes out-of-investigation cells and lack of cell equalization ability. Therefore, in the smart module concept, cells are first grouped into modules, which are then connected to the battery pack. Each module is an independent unit with a controller to investigate and control cells. Based on this concept, the work in this thesis redistributes tasks among module controllers and a central controller, applies a self-power design to enhance module independence, and selects the newly developed automotive ICs and sensors. Finally, a prototype of the BMS has been developed and successfully applied in a series of HEVs.
State of charge (SoC) is a battery state indicating its residual capacity. It is the fundamental state of the battery and is the basis for other battery operations. However, SoC is not a directly measurable state and has to be obtained by estimation techniques. Aiming to enhance the anti-noise ability of SoC estimation in a real vehicle environment, we propose a SoC estimation framework consisting of an adaptive nonlinear diffusion filter to reduce the noise of current measurement, a self-learning mechanism to remove its zero-drift, an open loop coulomb counting estimator and a model based closed loop filter to estimate SoC, and a data fusion unit to reach the final estimation result. In a simulation study, the closed loop filter is implemented based on an RC model and H∞ filter. In experiments and application, we modify the enhanced self-correcting model to model a type of LiFePO4 battery and apply an extended Kalman filter to estimate SoC. The framework has been demonstrated to improve accuracy and anti-noise ability, and achieves the technique upgrading goal recently published by the Chinese government.
Cell equalization is a crucial technique to balance the cells inside a battery pack, with the ability to maximize pack capacity and protect cells from damage. For the bi-directional Cuk equalizing circuit, we propose a SoC based, instead of voltage based, fuzzy controller to intelligently determine the equalizing current, with the aim of reducing equalizing duration, enhancing equalizing efficiency, and protecting cells. The inputs to the controller are specially designed as the difference in SoC, the average SoC, and the total internal resistance. Because of the lack of theoretical analysis on equalizing current in the electrochemistry field, we utilize a fuzzy controller to incorporate the experience and knowledge of experts. Simulations and experiments verify its availability and efficacy. Especially for a LiFePO4 battery, a large SoC difference may lead to only a small difference in voltage and cause the failure of a traditional voltage based equalizer. The SoC based method successfully avoids this problem and obtains good performance in equalizing LiFePO4 cells.
Fast charge is intended to charge a battery as fast as possible, without any damage and with high energy efficiency, thus helping to reduce vehicle out-of-service time and promote the commercialization of EVs. Battery safety and charging efficiency are partially reflected by the increase in temperature during the charging process. Therefore, the aims of this thesis were to accelerate charging speed and reduce the temperature increase. We introduce a model predictive control framework to control the charging process. An RC model and the modified enhanced self-correcting model are employed to predict the future SoC in simulations and experiments respectively. A single-node lumped-parameter thermal model and a neural network trained by real experimental data are also applied respectively. In addition, a genetic algorithm is applied to optimize the charging current under multiple objectives and constraints. Simulation and experimental results strongly demonstrate that the Pareto front of the proposed algorithm dominates that of the popular constant current constant voltage charge method.
با تشکر
ieeeproject
این وبلاگ توسط جمعی از دانشجویان ارشد و دکتری در زمینه مهندسی برق با هدف تبادل اطلاعات و فروش محصولات دانش بنیان با قیمتی مناسب راه اندازی شده است.در این وبلاگ مطالب زیر برای فروش قرار داده شده است .