Battery Power-Delivery Prediction Using Offset-Corrected Internal Resistance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing battery monitoring systems fail to accurately predict the time a battery can be used under specific load conditions and do not dynamically update for changes in battery behavior due to aging, leading to under-utilization and the need for frequent testing.
Innovation Solution
A method that retrieves battery operating parameters, calculates an offset between modeled and observed internal resistance, and updates the power-delivery performance prediction model to account for changes in battery behavior, providing accurate predictions of run time or distance to end-of-discharge based on user-defined inputs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If battery status indicators are based on state of charge or voltage, then the battery system provides basic status monitoring, but the system cannot determine the time for which the battery can be used under a particular load cycle
Solution Approach 1:
The system transitions from monitoring basic parameters (voltage, state of charge) to tracking dynamic parameters including internal resistance changes over time. By monitoring how internal resistance evolves during discharge cycles and comparing it to a degradation model, the system can predict remaining run time under specific load conditions, thus gaining time prediction capability without losing basic status monitoring
Solution Approach 2:
The patent replaces physical testing methods with a computational modeling approach. Instead of regularly performing physical load tests to assess battery health, the system uses a degradation model that correlates internal resistance changes with run time reduction, calculating predictions through mathematical relationships rather than mechanical testing
2Reliability
If the battery system is regularly tested to check whether it can still provide sufficient back-up run time, then run time accuracy can be maintained, but the complexity and frequency of testing increases
Solution Approach 1:
The system implements continuous feedback by monitoring internal resistance during normal operation and automatically updating the degradation model. This real-time feedback mechanism allows the system to maintain reliable run time predictions without requiring external testing interventions, as the model self-corrects based on observed resistance changes during each discharge cycle
Solution Approach 2:
The battery management system performs self-diagnosis and self-calibration by comparing actual discharge performance against the degradation model. The system automatically detects deviations between predicted and actual run times, then adjusts the model parameters accordingly, eliminating the need for external testing services while maintaining prediction accuracy
3Ease of manufacture
If a-priori testing is used to provide battery status indication, then initial battery characterization is available, but the system does not provide dynamic update of testing to account for variance in battery behavior due to aging
Solution Approach 1:
The system transforms the static a-priori testing approach into a dynamic model that evolves with battery aging. The degradation model is initially calibrated using manufacturer data, then continuously adapted during operation by incorporating real-time measurements of internal resistance and discharge performance, allowing the system to track and predict aging effects throughout the battery lifecycle
Solution Approach 2:
The patent performs preliminary model calibration using a-priori testing data from the manufacturer, establishing baseline parameters for the degradation model before the battery enters service. This preliminary action provides the foundation for subsequent dynamic updates, combining initial characterization with ongoing adaptive learning to maintain accuracy throughout the battery's operational life
Data Source
AI summary
Battery management systems and methods related to predicting power-delivery performance are provided. In one embodiment, a method for predicting power-delivery performance for a battery includes retrieving a plurality of battery operating parameters for a selected discharge cycle, calculating an offset indicative of a difference between a modeled internal resistance of the battery and an observed internal resistance generated from a calibration discharge cycle of the battery prior to the selected discharge cycle, and outputting a battery power-delivery performance prediction based on an offset-corrected internal resistance indicative of a difference between a modeled internal resistance based on the plurality of battery operating parameters for the selected discharge cycle and the offset.