Edge Battery Diagnosis Using Differential Convolution Thresholds
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing battery management systems are inefficient in providing real-time, rapid diagnosis of battery performance and health, particularly in complex energy storage scenarios, where lengthy data collection and analysis processes hinder immediate detection of underperforming batteries and safety issues.
Innovation Solution
A method utilizing edge computing to periodically collect battery data, perform convolution using first-order differential operators, and identify anomalies outside a threshold range, enabling rapid and precise battery diagnosis directly on edge devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If battery data is uploaded to a cloud-based platform for extensive data analysis and forecasting, then comprehensive battery health prediction is achieved, but the data collection and analysis process becomes lengthy and real-time performance deteriorates
Solution Approach 1:
The patent segments the battery diagnosis system into two parts: edge devices that perform rapid real-time monitoring and anomaly detection using convolution operations, and cloud platforms that handle comprehensive long-term analysis. This segmentation allows simultaneous achievement of real-time responsiveness and thorough analytical capabilities by distributing different diagnostic functions across different computational layers.
2Reliability
If extensive data analysis is performed on cloud-based platforms, then comprehensive battery forecasting is achieved, but the system complexity and computational burden increase
Solution Approach 1:
The patent applies preliminary action by performing initial data filtering, convolution operations, and anomaly detection at the edge device level before data is transmitted to the cloud. This preliminary processing reduces the complexity of subsequent cloud-based analysis by pre-processing data locally and only transmitting relevant anomaly information or aggregated statistics to the cloud platform.
3Speed
If real-time battery monitoring is implemented, then immediate anomaly detection is achieved, but the computational resources required at edge devices increase
Solution Approach 1:
The patent implements partial action by applying convolution operations selectively to detect specific anomaly patterns rather than performing exhaustive analysis on all battery data parameters. The system uses first-order differential operators and threshold-based anomaly detection that require minimal computational resources while still achieving rapid real-time monitoring capabilities at edge devices.
Data Source
AI summary
A method for rapid battery diagnosis based on edge computing, storage medium and electronic device. The method includes: periodically collecting battery information data from a battery management system at a preset frequency; performing convolution on the battery information data using first-order differential operators to obtain primary convolution results; identifying secondary convolution results that fall outside an underperforming threshold range based on the primary convolution results; and detecting anomalies in the battery information data based on the secondary convolution results and the battery information data. The present disclosed method allows for more precise and rapid processing of battery data on edge devices, enables a quick diagnosis of battery performance, and is easily scalable.