Big Data Battery Diagnostic System for Voltage Estimation Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional battery cell safety diagnostic methods are passive and based on physical formulas, failing to accurately account for temperature variations and bus bar resistance, leading to inconsistencies in battery state estimation and poor accuracy in mass-produced vehicles.
Innovation Solution
A big-data-based battery diagnostic system that collects and learns from data from hundreds to tens of thousands of mass-produced vehicles to set a reference voltage for each vehicle, allowing for real-time comparison and analysis of battery state information to diagnose deviations and provide customized safety diagnoses.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If conventional formula-based battery cell safety diagnostic method is used, then the diagnostic system is simple to implement, but the accuracy and consistency of battery state estimation deteriorate due to inability to account for temperature variations and bus bar resistance
Solution Approach 1:
The patent introduces a big data server as an intermediary between the battery management system and the diagnostic algorithm. This server collects, stores, and processes battery data from multiple vehicles, providing refined reference data that accounts for temperature variations, bus bar resistance, and other environmental factors. This resolves the contradiction by enabling high-accuracy diagnostics without requiring complex local computational resources in each vehicle's BMS.
Solution Approach 2:
The patent transitions from single-vehicle isolated diagnostics to multi-vehicle collaborative diagnostics by adding the dimension of fleet-wide data aggregation. Instead of diagnosing one vehicle in isolation with limited data, the system aggregates data from hundreds to thousands of vehicles across different environments, transforming the diagnostic accuracy from a single-point measurement to a multi-dimensional statistical analysis that accounts for various operating conditions.
2Temperature
If battery modeling is used to control battery temperature, then the system can manage thermal conditions, but the electrochemical characteristics vary greatly due to temperature, surrounding environment, and cooling systems making accurate control difficult
Solution Approach 1:
The patent implements a feedback mechanism where actual battery data from the vehicle is continuously compared against reference data derived from big data analysis. The system uses this feedback to dynamically adjust diagnostic thresholds and parameters, accounting for temperature variations, cooling system effects, and environmental conditions. This resolves the reliability issue by continuously adapting to actual operating conditions rather than relying on static models.
Solution Approach 2:
The patent dynamically changes diagnostic parameters based on operating conditions. Instead of using fixed thresholds for battery state estimation, the system adjusts voltage, temperature, and current thresholds based on real-time environmental conditions, state of charge, and historical data from similar operating scenarios. This allows accurate diagnostics across varying temperature and environmental conditions.
3Measurement precision
If conventional battery modeling methodology is used, then voltage estimation at room temperature can be performed, but the method cannot reflect actual state of bus bar resistance, electric load, and BSA
Solution Approach 1:
The patent performs preliminary analysis of battery characteristics across diverse operating conditions by aggregating data from multiple vehicles before deployment. The big data server pre-processes and refines reference data for various temperature ranges, load conditions, and state of charge levels. This preliminary action enables the system to quickly adapt to new operating conditions without requiring extensive real-time computation or retraining.
4Measurement precision
If testing and modeling of all battery cells is performed, then individual cell characteristics can be captured, but mass production becomes infeasible due to the large number of battery cells in each vehicle
Solution Approach 1:
The patent creates a virtual model of battery behavior by copying and aggregating data from multiple actual vehicles. Instead of physically testing each cell, the system uses big data to create a statistical representation of battery characteristics that captures individual cell variations through fleet-wide analysis. This virtual copying approach enables individual cell-level diagnostics without the need for physical testing of every cell during manufacturing.
Data Source
AI summary
Disclosed herein is a big-data-based battery diagnostic system that includes a first vehicle having a first battery module disposed therein, a battery data server having a processor configured to collect first state information related to the first battery module provided from the first vehicle, to learn the collected first state information, and to set a reference voltage for the first vehicle based on the learned first state information, and a second vehicle configured to receive the reference voltage from the battery data server and having a second battery module disposed therein. The second vehicle is configured to measure second state information related to the second battery module, compare the measured second state information with the reference voltage to analyze the same, and apply a result value of the comparison and analysis to a preset diagnostic level range to diagnose a current state of the second vehicle.