.png){kind=icon}
.png){kind=icon}
Return
Battery Temperature Forecasting Method: Li-Ion Batteries Case Study
.png){kind=icon}
PRE
AI
DOI：10.1007/978-981-97-3556-3_5.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Monitoring and managing battery health is crucial for enhancing performance and lowering running expenses for electronic devices. Our study explores AI-powered temperature forecasting models specific to lithium-ion battery types, in instances where these batteries have been tested independently. This research presents time series forecasting approaches to predict temperature for the battery packs. We propose autoregressive integrated moving average (ARIMA) and long short-term memory (LSTM) for predicting the battery temperature and beware of probable future temperatures beforehand to minimize the chances of overcharging and prevent the battery from crossing the threshold value above which battery's health characteristics might get hampered. With the increase in adoption of data-directed approaches for battery forecasting, we illustrate the competence of ARIMA and LSTM in conditions where there are hardly any preceding details obtainable about the batteries. With respect to this task, we possess a distinct dataset of 34 lithium-ion battery units. In one respect, outcomes suggest that the established ARIMA model supplied pertinent ways to interpret the information through a variety of battery types. Having said that, LSTM model outcomes recommend that the developed univariate and multivariate LSTM model provides finer prediction exactness provided that we have a greater diversification in data available for a given battery type. We thus try to generalize one forecasting model for each battery type depending on the model's performance.
Keywords:
Lithium-ion batteries
ARIMA
Machine learning
LSTM
Time series
Journal
P
IF:
0
Papers:
9
Citations:
0