Battery State Prediction Using Charge-Discharge Data Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing battery management systems face efficiency degradation due to managing large-volume battery data, leading to reduced performance and accuracy of artificial intelligence models used for predicting battery state, necessitating improved data selection for training.
Innovation Solution
A battery state prediction apparatus extracts specific battery data subsets for training deep learning models, using first and second battery data after charging and discharging phases, and applies these to separate CNN models to generate combined state data for accurate battery state prediction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If large-volume battery data is used for training the artificial intelligence model, then the model has access to more comprehensive information, but the system undergoes efficiency degradation and performance reduction
Solution Approach 1:
The patent extracts specific critical features from large-volume battery data including voltage, current, temperature, and state of charge information. By selecting and extracting only the most relevant features rather than processing all raw data, the system maintains prediction accuracy while significantly reducing computational burden and improving processing efficiency.
Solution Approach 2:
The patent creates a simplified representation or copy of the essential battery state information through feature extraction. Instead of processing the complete large-volume raw data, the system works with extracted feature vectors that capture the critical aspects needed for accurate state of health prediction, thereby improving efficiency without sacrificing reliability.
2Measurement precision
If comprehensive battery data is processed, then more complete battery state information is obtained, but the processing time and computational resources increase
Solution Approach 1:
The patent extracts essential features from comprehensive battery data including voltage, current, temperature, and state of charge. By taking out only the critical features needed for accurate state prediction rather than processing all comprehensive data, the system reduces processing time while maintaining measurement precision.
Solution Approach 2:
The patent applies partial action by processing only the necessary subset of battery data features rather than all available comprehensive data. This selective processing of essential features (voltage, current, temperature, SOC) achieves accurate state prediction without the time cost of processing excessive comprehensive information.
3Reliability
If all battery data is analyzed by the artificial intelligence model, then complete battery state assessment is achieved, but the model performance and speed are reduced
Solution Approach 1:
The patent extracts key features from all battery data including voltage, current, temperature, and state of charge before feeding them to the artificial intelligence model. By taking out only the essential features rather than analyzing all raw battery data, the system maintains prediction reliability while significantly improving model inference speed.
Solution Approach 2:
The patent applies partial action by providing only the necessary feature subset to the AI model rather than all complete battery data. This selective input of essential features (voltage, current, temperature, SOC) achieves reliable battery state assessment while maintaining high model inference speed.
Data Source
AI summary
Discussed is a state prediction apparatus that may include a data managing unit configured to extract first battery data including battery data obtained for a first predetermined time after completion of charging of a battery and second battery data including battery data obtained for a second predetermined time after entering of discharging of the battery and a controller configured to obtain first state data for predicting a state of the battery by applying the first battery data to a first deep learning model, obtain second state data for predicting the state of the battery by applying the second battery data to a second deep learning model, and predict the state of the battery based on the first state data and the second state data.