Battery-Degradation Model Based on ANN-Regression Function for EV Applications

Document Type

Conference Proceeding

Publication Date


Publication Title

Proceedings of the Global Humanitarian Technology Conference






Lithium-ion batteries are currently the most widely used form of energy storage in electric vehicles. They have a high vitality thickness with a potential for higher limits, they don't need prolonged priming when they're new. However, they're subject to aging rapidly even when they aren't being used. Being able to predict the life expectancy of these batteries can prove to be beneficial in order to see the limitations of using them in vehicles. However, continuous use of these batteries eventually leads to a shorter and shorter battery life. The goal of this work is to implement ANN (Artificial Neural Network) Prediction Model to connect between various characteristics of this battery type. ANN-Predictive model is implemented, trained, and tested based on empirical samples, improved Thevenin model, and MATLAB codes. This neural model can anticipate values in - between learning values, likewise make introduction between expectations to learn and adapt information at different qualities. Arithmetical nonlinear capacities which, interfaces amongst information sources and yields for neural systems with its related Simulink model are concluded. This is done keeping in mind the end goal to help any scientist without the need of preparing the neural system each time. This model' sources of information are: the Time and SOC (State-of-Charging). Its yields are: Average Degradation Function (ADF), Degradation Density Function (DDF), Cycle Life L(x), Depth of Discharge (DoD) and Capacity rate. It contains two layers, one covered up with log-sig capacity and 10 neurons; and second layer has pure-line work with 5 neurons. The electric vehicle technology and its associated storage devices can help the humanitarian to go beyond cutting edge technology and help small villages. One day, the electric vehicles will be the next cell phone where everyone will have one. Final presentation includes characteristics, regression, comparisons and Lithium-ion battery ANN curves.