Machine Learning Battery Fault Detection from Voltage and Temperature
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Rechargeable batteries, particularly in electric vehicles, are prone to faults that can cause operational disruptions and safety hazards, and existing detection methods lack real-time accuracy and consider variations in cell chemistries and applications.
Innovation Solution
A machine learning-enabled system that utilizes multiple sensors and AI models to detect faults at the cell, module, and pack levels by analyzing voltage and temperature data, with cloud-based model updates for enhanced accuracy and real-time detection, enabling preventative actions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional fault detection methods are used for battery packs, then the system complexity is low, but the measurement precision and reliability of fault detection are insufficient
Solution Approach 1:
The patent replaces traditional mechanical/electrical threshold-based detection systems with a machine learning-based detection system. The machine learning model processes voltage and temperature measurements to predict fault conditions, achieving higher measurement precision while managing system complexity through software-based intelligence rather than complex hardware circuits.
Solution Approach 2:
The patent changes the detection approach from fixed threshold parameters to dynamic parameters derived from machine learning models. The system uses voltage scores and temperature scores that are dynamically calculated based on learned patterns from training data, allowing the detection criteria to adapt to different battery conditions and chemistries.
2Measurement precision
If machine learning models are trained individually for each cell chemistry and application, then the measurement precision improves, but the device complexity and training time increase significantly
Solution Approach 1:
The patent creates a universal machine learning model that can detect faults across different cell chemistries and applications. The model is trained on diverse data from multiple sources and is designed to generalize to new battery types without requiring separate training for each chemistry or application, thereby reducing overall system complexity while maintaining high detection accuracy.
Solution Approach 2:
The patent performs preliminary training actions by pre-training the machine learning model on extensive diverse data before deployment. This preliminary training enables the model to learn general fault patterns across different battery types, reducing the need for extensive retraining when deployed with different cell chemistries or applications.
3Measurement precision
If comprehensive voltage and temperature measurements are collected from multiple sensors, then the measurement precision improves, but the loss of time for data processing increases
Solution Approach 1:
The patent extracts only the essential features from comprehensive sensor measurements by calculating voltage scores and temperature scores that capture the most relevant information for fault detection. This extraction process reduces the dimensionality of the data while preserving the critical fault indicators, enabling faster processing without sacrificing detection accuracy.
Solution Approach 2:
The patent uses partial action by selecting and processing only the most critical voltage and temperature measurements needed for fault detection, rather than processing all available sensor data equally. The machine learning model is designed to focus on the most informative features, reducing processing time while maintaining high detection precision.
Data Source
AI summary
In one aspect, computer-implemented method may include receiving, from one or more sensors associated with a battery pack, one or more measurements pertaining to voltage, temperature, or both. The method may include determining, based on the one or more measurements, a voltage score and a temperature score, and predicting, based on the voltage score and the temperature score, whether the battery pack is experiencing a fault condition. The prediction is performed by an artificial intelligence engine. Responsive to predicting the battery pack is experiencing the fault condition, the method may include performing one or more preventative actions.