Battery State Estimation With Confidence-Scored Predictive Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Battery management systems face challenges in accurately determining state of charge (SOC) and state of health (SOH due to variations in environmental conditions and usage scenarios, leading to inaccuracies across different battery types and chemistries.
Innovation Solution
A predictive model implementing a recursive algorithm generates confidence scores to correct and update SOC and SOH measurements by considering battery-specific parameters, using feature, chemistry, and system matrices to improve accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional battery management systems are used to determine SOC and SOH, then the system is simple and easy to implement, but the measurement precision is insufficient due to variations in environmental conditions and usage scenarios
Solution Approach 1:
The patent segments the battery management system into multiple specialized modules: a confidence scoring module that evaluates measurement reliability, a correction module that adjusts SOC/SOH values based on confidence scores, and a database module that stores battery-specific parameters. This segmentation allows each module to perform its specific function with high precision while maintaining overall system manageability.
Solution Approach 2:
The system performs preliminary actions by pre-storing battery-specific parameters, confidence score thresholds, and correction factors in a database during the design phase. These pre-computed values are readily available when needed, enabling rapid and accurate SOC/SOH determination without complex real-time calculations, thus improving measurement precision without proportionally increasing operational complexity.
2Measurement precision
If battery management systems account for different environmental conditions and usage scenarios, then the measurement precision improves, but the device complexity increases due to need for multiple correction factors and algorithms
Solution Approach 1:
The patent changes parameters by introducing confidence scores as an additional dimension for evaluating measurement quality. Instead of using fixed correction algorithms, the system dynamically adjusts the weighting and selection of correction factors based on the calculated confidence score, which reflects the reliability of measurements under specific environmental conditions and usage scenarios. This parameter-based approach improves precision across varying conditions while maintaining algorithmic simplicity.
Solution Approach 2:
The system implements feedback through the confidence scoring mechanism, where the confidence score generated from real-time measurements feeds back into the correction process. The confidence score indicates the reliability of current measurements, and this feedback loop allows the system to automatically adjust correction applications without complex decision logic, thereby improving measurement precision while keeping the control algorithm straightforward.
Data Source
AI summary
Methods and systems are provided for using a predictive model to monitor batteries and other energy devices. In some examples, the predictive model may receive, as input, data indicating one or more parameters of a battery system. In such examples, the predictive model may generate, as output, one or more corrections or updates to the one or more parameters. In certain examples, the one or more parameters may include a state-of-charge of the battery system and/or a state-of-health of the battery system. In certain examples, the one or more corrections or updates may include a confidence score corresponding to an accuracy of the state-of-charge and/or an accuracy of the state-of-health. In some examples, the one or more corrections or updates may be updated, according to one or more outputs of the predictive model, responsive to one or more convergence criteria not being met.