Battery Aging Prediction Using Fleet Usage Profiles
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Accurately determining the aging state of a battery is challenging due to various influencing factors, such as current throughput, charging cycles, temperature, and state of charge, which individually affect the state of health (SOH) value, making it difficult to predict battery aging effectively.
Innovation Solution
A method involving data collection from lithium-ion batteries, including voltage, current, and temperature patterns, which are processed using a data-driven fleet model and battery reference model, incorporating machine learning and AI to predict aging states by analyzing usage profiles and comparing with stored data from comparison batteries, allowing for extrapolation and association of aging patterns to determine predicted aging states and patterns.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple influencing factors (current throughput, charging cycles, temperature, state of charge) are considered individually, then the complexity of determining aging state increases, but the accuracy of aging state determination deteriorates
Solution Approach 1:
The patent combines multiple individual influencing factors (current throughput, charging cycles, temperature, state of charge) into a unified data-driven fleet model that processes all factors simultaneously. This integration allows the system to capture complex interactions between factors while maintaining determination accuracy, resolving the contradiction between considering multiple factors and system complexity.
Solution Approach 2:
The fleet model serves multiple functions: it processes various usage profiles, handles different battery types, performs aging state determination, and enables comparison across multiple batteries. This multi-functionality allows a single system to handle diverse inputs and produce accurate aging state predictions without requiring separate determination systems for each factor.
2Measurement precision
If data from multiple comparison batteries are processed and analyzed, then the accuracy of aging state prediction improves, but the computational complexity and data processing requirements increase
Solution Approach 1:
The system performs preliminary processing of usage profile data from multiple comparison batteries before aging state determination. By pre-processing and organizing data from multiple sources in advance, the fleet model can efficiently compare and analyze the data without overwhelming computational complexity during the actual aging state prediction process.
Solution Approach 2:
The patent uses comparison batteries as proxies or copies to represent aging patterns. Instead of directly measuring every parameter of the target battery, the system creates aging state predictions by comparing against patterns from multiple comparison batteries, reducing the computational burden while maintaining prediction accuracy.
3Reliability
If usage profiles and battery-characterizing variables are continuously monitored and analyzed, then the reliability of aging state determination improves, but the loss of information and processing overhead increase
Solution Approach 1:
The fleet model extracts only the essential battery-characterizing variables and usage profile features needed for aging state determination, rather than processing all available data. This extraction of critical information maintains determination reliability while reducing processing overhead and preventing information loss from excessive data handling.
Data Source
AI summary
A method for predicting an aging state of a battery. A vehicle is also provided that includes at least one battery, the aging state of which is predited using the method and/or the aging behavior of which is improved based on the aging state prediction. A prediction system that is configured to carry out the method is also provided.