Corrected Sample Entropy for Battery Pack Fault Diagnosis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current entropy-based methods for battery fault diagnosis in lithium-ion battery packs are inadequate in accurately and quickly detecting fault types and times, and are prone to high computational costs and poor robustness.
Innovation Solution
A multi-fault diagnosis method and system based on corrected sample entropy, which involves measuring cell voltages, constructing a cell voltage sequence, calculating sample entropy, applying a correction coefficient to represent voltage fluctuations, and judging fault types based on changes in the corrected sample entropy values.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If current entropy-based methods are used for battery fault diagnosis, then the method is easy to implement and has low computational cost, but it cannot accurately and quickly detect early fault types and time
Solution Approach 1:
The patent transforms the raw voltage sequence into corrected sample entropy values by introducing a correction coefficient that adjusts for voltage fluctuations. This parameter transformation enables the system to distinguish between normal voltage variations and actual fault conditions, thereby improving detection accuracy while maintaining computational efficiency. The corrected sample entropy value serves as a new diagnostic parameter that captures both the temporal and amplitude characteristics of voltage behavior.
Solution Approach 2:
The correction coefficient acts as an intermediary element between the raw voltage measurements and the fault diagnosis conclusion. It mediates the relationship by normalizing voltage fluctuations against a reference profile, allowing the system to detect early faults before they manifest as obvious voltage deviations. This intermediary transformation enables earlier and more accurate fault detection without requiring complex modeling.
2Reliability
If current entropy-based methods are used for battery fault diagnosis, then the implementation is simple, but the robustness is poor and computational cost is high
Solution Approach 1:
The patent modifies the sample entropy calculation by introducing a correction coefficient that accounts for voltage fluctuations. This parameter adjustment enhances the robustness of the diagnostic method against noise and voltage variations while keeping the computational structure relatively simple. The corrected sample entropy value provides a more reliable indicator of fault conditions without requiring complex computational procedures.
3Measurement precision
If a correction coefficient is introduced to represent voltage fluctuation information, then fault detection accuracy is improved, but the computational steps increase
Solution Approach 1:
The correction coefficient is derived from the voltage sequence itself through a systematic transformation process. By defining the correction coefficient based on the relationship between current voltage and reference voltage profiles, the patent achieves improved fault detection accuracy without introducing external complexity. The additional computational step is a straightforward comparison and scaling operation that enhances diagnostic precision.
Data Source
AI summary
A multi-fault diagnosis method has the following steps: measuring cell voltages of a battery pack to be diagnosed; constructing a cell voltage sequence according to measured cell voltages of the battery pack to be diagnosed, and calculating a sample entropy value of the cell voltage sequence; setting a correction coefficient for representing voltage fluctuation information, and correcting the sample entropy value through the correction coefficient to obtain a corrected sample entropy value; and judging and outputting a fault type of the battery pack to be diagnosed according to a numerical value change of the corrected sample entropy value. Faults of cells can be accurately diagnosed without a model, sample entropy values under different faults can be distinguished by setting the correction coefficient, the intuitiveness and efficiency of fault detection are improved, and the fault type and time of the lithium-ion cells can be quickly, accurately and stably diagnosed and predicted.