الفهرس 
INTRODUCTIONOnly 14 pages are availabe for public view
 |  | from | 191 |  |  |
| | | from | 191 | | |
Abstract
Crude oil, ethanol, petrol, and diesel fossil fuel sources are rapidly running out every day because of the rising demand for transportation. Since their costs are increasing as a result, alternative fuel vehicles like electric vehicles are becoming more popular. For storing energy, Batteries are utilized as Energy Storage Systems (ESS) to store the energy produced by alternative fuel vehicles. To ensure the battery’s safe and dependable functioning, it must be precisely modeled. Additionally, the state of charge (SoC) must be calculated to prevent overcharging and over-discharging, which can harm the battery’s internal structure.
The primary goals of the thesis are to develop an accurate state of charge estimate and parameter modeling for a lithium-ion battery. Since parameter identification is a nonlinear optimization process problem, the recommended identification method is carried out utilizing a precise state space model built from an analogous electric circuit model. When compared to other optimizers like the Nelder-Mead Simplex algorithm, Quasi-Newton algorithm, Runge Kutta Optimization algorithm (RK), Genetic algorithm (GA), Grey Wolf Optimization algorithm (GWO), and Gorilla Troops Optimization algorithm (GTO), the African Vultures Optimization Algorithm (AVOA) is used to identify the battery characteristics and solve the problem. Contrary to earlier hybrid approaches, such as the Coulomb counting method with Extended Kalman Filter (EKF), the Coulomb counting method with Unscented Kalman Filter (UKF), and the Coulomb counting method with Adaptive Extended Kalman Filter (AEKF), an efficient hybrid method combining the Coulomb counting method (CCM) and Adaptive Unscented Kalman Filter (AUKF) is used to precisely estimate the SoC of lithium-ion batteries.
To identify the battery properties and estimate its SoC at various states, which vary between incorporating loads and battery fading effect or not under varied temperatures, four scenarios are applied to the battery. To illustrate the usefulness of AVOA and hybrid approaches for parameter identification and SoC estimate of the battery, these numerical simulations are carried out on a 2600mAhr Panasonic rechargeable lithium-ion battery. In these cases, the fitness function is defined by the integral of the estimated and measured voltage as well as the estimated and measured state of charge.
The investigation and analysis of simulation results reveal that, in comparison to other algorithms, the suggested AVOA for parameter identification of the battery and the hybrid method of CCM and AUKF are carried out with great precision, least error, and great proximity to the experimental data. As a result, it is easier and more practical to implement adaptive algorithms and procedures on a lithium-ion battery without having to worry about overcharging or over-discharging the battery.