State-of-Health Prediction

• Tech Stack: Seaborn, Scipy, Pandas, Tensorflow

• Members: Me, Carlos Rufino

Health status is a signal indicating the percentage value of degradation through certain metrics.

From the discharge capacity, we can calculate the health status of batteries that are about to fail or enter a second life.

Bidirectional-LSTM memory units can capture vectors from a feature mapping derived from the convolutional filter.

The neural network model mentioned is responsible for performing the forecasting steps, i.e., predicting how many cycles each battery will last depending on the temperature, current, voltage, and dQdV.