Predicting the State of Health of Batteries with Machine Learning
General description of the project
This research focuses on applying machine learning (ML) to evaluate the State of Health (SOH) of re-chargeable batteries, which are widely used in renewable energy related applications. Battery test data will be gathered and prepared to train a ML model. Using Tensorflow and Keras which cost are virtually free.
Technologies
Initial utilization of Digital Multimeters to record battery voltage. Eventually moved onto tests utilizing Rigol Oscilloscope as a load to measure battery capacity and discharging curves of certain Lithium-ion batteries. The data used was information on battery cells used in a Flying Electric Vehicle that has recorded voltage, current and temperature. The goal was to create a Machine Learning Model that can predict the state of health in Lithium-ion Batteries with the data given in the excel worksheets we gathered.
Explain project results
As a Hispanic, my institution and researching in the Department of Education (DOED) Mentored Research has allowed for the facilitation of my learning in this field. With the support that is given by Scholars Academy and its connections allows for us Hispanics to advance further in our careers.
Why it should be considered best practice?
This project is a “Best Practice” due to how any student who desires to study Machine Learning and the field of control and instrumentation engineering technology can replicate the process of researching battery tests and creating a machine learning model. All that is needed is dedication and a computer!
Highlights of your proposed presentation
Learning Machine Learning Models, how batteries are used in everyday applications, and the importance of measuring these batteries state of health.
The Evaluation Committee will evaluate submitted proposals based on the following criteria. Each area will be rated on a scale from 1 to 7 (1= non-satisfactory; 7 =outstanding), for a maximum of 63 points.