Battery Error Prediction Using Risk Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for predicting errors in device batteries often require immediate user intervention without adequately distinguishing between different error types, leading to unnecessary repairs or safety hazards.
Innovation Solution
A computer-implemented method using an error evaluation model with an error factor assignment table to detect anomalies in device battery operational variables, evaluate error-relevant features, and calculate a risk value based on propagation speed, severity, and probability, thereby providing tailored instructions for action.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If rule-based evaluation with threshold value comparisons is used to detect abnormalities, then early error detection is enabled, but all abnormalities are treated equally requiring immediate user intervention
Solution Approach 1:
The patent segments abnormalities into different error types (incipient errors, critical errors, hazardous errors) based on their characteristics and risk levels. This segmentation allows differentiated user instructions rather than treating all abnormalities equally, resolving the contradiction between reliable detection and ease of operation.
Solution Approach 2:
The patent changes parameters by evaluating multiple error-relevant variables (temperature behavior, state of charge profile, aging state, charging behavior, cell pressure) and their gradients to classify error types. This parameter-based classification enables differentiated responses based on the severity and nature of each abnormality.
2Measurement precision
If multiple error-relevant variables are monitored with higher sampling rates, then error classification accuracy is improved, but system complexity increases
Solution Approach 1:
The patent extracts only the most relevant error indicators from the monitored variables, such as gradients of error-relevant variables and specific histograms. This extraction approach maintains high classification accuracy while reducing the complexity of the monitoring system by focusing on key parameters rather than processing all variables equally.
Solution Approach 2:
The patent monitors multiple error-relevant variables with higher sampling rates only when anomalies are detected, rather than continuously monitoring all variables at high resolution. This partial action approach improves error classification accuracy when needed while reducing overall system complexity and resource consumption.
3Loss of time
If predictive error detection methods are implemented, then user notification time is extended, but false predictions may lead to unnecessary repairs
Solution Approach 1:
The patent uses feedback by continuously monitoring error-relevant variables and updating the error type classification as new data becomes available. This feedback mechanism allows the system to confirm or revise predictions, reducing false positives while maintaining extended user notification time for incipient errors that are likely to develop into critical failures.
Solution Approach 2:
The patent performs preliminary evaluation of multiple error-relevant variables and their trends before making final error predictions. This preliminary action includes analyzing gradients and patterns that indicate developing failures, allowing extended notification time for genuine risks while filtering out false predictions through thorough preliminary assessment.
Data Source
AI summary
A computer-implemented method for providing a risk value for a predicted error in a device battery of a technical device using an error evaluation model, wherein the error evaluation model has at least one error factor assignment table. In one example, the method includes detecting temporal operational variable profiles of at least one device battery; performing an anomaly detection as a function of the temporal operational variable profiles; upon recognizing an anomaly, detecting error-relevant variables; evaluating the error evaluation model as a function of the error-relevant variables in order to determine an error type of a predicted error; assigning error factors to the error type using the at least one provided error factor assignment table of the error evaluation model; determining a risk value as a function of the error factors; and signaling the risk value.