Battery Sensor Fault Diagnosis Using Zonotope Kalman Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing fault diagnosis methods for power battery management systems cannot accurately detect sensor faults, leading to potential safety hazards such as over-charging or over-discharging due to incorrect battery state estimation.
Innovation Solution
An uncertain noisy filtering-based fault diagnosis method is introduced, which establishes a second-order Thevenin equivalent circuit model and an electro-thermal coupling model to expand state constraints into system output vectors, using zonotope sets and strip spaces to estimate sensor faults in core and surface temperature sensors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If traditional state estimation method is used to diagnose battery faults, then the diagnosis process is simple, but the diagnosis accuracy deteriorates when sensor faults occur
Solution Approach 1:
The patent introduces an uncertain noisy filtering method as an intermediary layer between sensor data collection and fault diagnosis. This filtering mechanism processes sensor signals before they are used for state estimation, thereby mediating the impact of sensor faults on diagnosis accuracy while maintaining a relatively simple overall diagnostic framework.
Solution Approach 2:
The patent changes the parameters of the estimation process by incorporating uncertainty modeling and noise filtering parameters. Instead of using fixed threshold values for fault detection, the system dynamically adjusts estimation parameters based on signal quality and uncertainty levels, improving diagnosis accuracy under varying sensor conditions.
2Speed
If sensor data is used directly for state estimation without filtering, then the processing speed is fast, but the reliability deteriorates when sensor faults are present
Solution Approach 1:
The patent applies preliminary uncertain noisy filtering to sensor data before the main state estimation process. This preliminary action removes or attenuates noise and fault effects from sensor signals in advance, ensuring that the subsequent fast state estimation operates on cleaned data, thus maintaining both speed and reliability.
Solution Approach 2:
The uncertain filtering mechanism acts as a cushioning layer that absorbs and mitigates the harmful effects of sensor faults before they can significantly impact the state estimation. This beforehand cushioning protects the reliability of the estimation process while allowing rapid processing to continue.
3Ease of operation
If fault diagnosis is performed without considering sensor fault possibilities, then the system operation is simple, but safety hazards occur due to incorrect state judgment
Solution Approach 1:
The patent implements a self-service fault detection mechanism where the uncertain filtering system automatically identifies and flags potential sensor faults during the normal state estimation process. This self-service approach detects sensor anomalies without requiring separate complex diagnostic procedures, maintaining operational simplicity while preventing safety hazards.
Data Source
AI summary
An uncertain noisy filtering-based fault diagnosis method for a power battery management system is described. The method includes establishing an electro-thermal coupling model of the power battery system; extending an output vector of the system according to a state constraint of a power battery, and expanding a state vector of the system according to a fault of the power battery system to obtain an augmented system of the power battery system; obtaining an estimation interval of a power battery sensor fault by using a zonotope Kalman filtering method; judging whether the power battery management system has a fault according to upper and lower bounds of fault estimation; if a fault occurs, determining a fault type and a fault time according to a result.