Battery Safety Determination via Feature Space Projection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing battery management systems face challenges in accurately determining battery safety due to dynamic changes in physical quantities like voltage and temperature, leading to potential sensor errors and unreliable abnormality detection, especially when batteries are in use.
Innovation Solution
A battery management method that acquires physical quantity data, calculates unbalance data, projects it to a feature space using a feature extraction model, and determines battery safety based on distribution information, allowing for real-time monitoring and alerting of abnormal states through a feedback system.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Use of energy by moving object
If battery capacity is increased to meet higher power consumption demands, then energy supply capability is improved, but safety risk increases due to greater potential damage from explosions or failures
Solution Approach 1:
The system performs preliminary actions by continuously monitoring physical quantities (voltage, temperature, current) and calculating unbalance data before actual battery failures occur. The feature extraction model projects unbalance data to feature space and compares it against normal distribution patterns, enabling early detection of abnormal states and preventive measures against battery explosions or failures.
2Measurement precision
If physical quantity data is monitored in real-time to improve safety detection, then measurement accuracy is improved, but sensor errors and unreliable detection increase due to dynamic changes during battery use
Solution Approach 1:
The system applies parameter changes by transforming raw physical quantity data into unbalance data, then projecting to feature space using a feature extraction model. This multi-stage transformation converts dynamic, noisy sensor readings into stable feature representations that can be reliably compared against normal distribution patterns, filtering out sensor errors and dynamic variations.
Solution Approach 2:
The feature extraction model acts as an intermediary between raw sensor data and safety determination. By projecting unbalance data to feature space and comparing against pre-established normal distribution patterns, the system mediates the relationship between dynamic physical quantities and safety assessment, reducing the impact of sensor errors and dynamic changes during battery use.
Data Source
AI summary
Provided is a battery management method and apparatus. The battery management method includes acquiring physical quantity data, for each of a plurality of batteries, of when corresponding physical quantities of the plurality of batteries, making up the physical quantity data, dynamically vary, calculating unbalance data based on physical quantity difference information derived from the physical quantity data, calculating feature data for the physical quantity data by projecting the unbalance data to a feature space, and determining a battery safety for one or more of the plurality of batteries based on determined distribution information of the feature data.