Abstract
The lithium-ion battery (LiB) has become increasingly popular in electric vehicles (EVs), laptops, phones and many other devices that people use every day. It is popular due to its high energy density, low cost, lack of memory effect (typical of older battery types), longer life cycle (more full cycles till battery dies), and more. Because of its common use in everyday applications, knowing the state of charge (SOC) of a lithium-ion battery becomes an important problem to solve.
The first goal of this work was to develop and present a new battery model derived from fundamental principles of electrochemistry, Fick’s first law of diffusion, and a mass balance of lithium-ions between the anode and cathode. A voltage equation was developed based on the the open circuit voltage, the Nernst equation, and Ohm’s law. The battery model aims to accurately track the movement of lithium ions without being too computationally demanding, while the voltage equation relates the output of the model to the voltage of the cell. These equations are coupled and solved simultaneously. The SOC of the battery could then be determined based on the mass of lithium in the anode.
The second goal of this work was to optimize the parameters of the battery model to make it match experimental data. A cost function was defined and a genetic algorithm was implemented to minimize the cost function by altering the model parameters. The genetic algorithm successfully reduced the cost function. The model matched the experimental data and is therefore ready for commercial application.
