Wireless Battery Management Off-Board Model Mapping

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current battery management systems for electric vehicles face limitations due to limited data storage and computing power, which restricts the use of historic data and data from other batteries, leading to suboptimal performance in estimating battery status and life.

Innovation Solution

A wireless network-based battery management system with an off-board subsystem for data storage and processing, allowing for the utilization of historic data and data from other batteries to establish accurate complex models, which are then mapped into simple models for on-board calculation, reducing the computational load and cost of on-board chips.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If complex battery models with advanced algorithms are used for accurate SOC estimation, then measurement precision is improved, but device complexity and computing power requirements increase

Engineering Contradiction:
ImproveSOC estimation accuracyVSAvoidcomputing power requirement
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The system divides the battery management functionality into two segments: complex model training and validation are performed offline on a server, while only simple model inference is executed online on the embedded controller. This segmentation allows accurate SOC estimation without requiring the embedded system to have high computing power.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system performs preliminary action by training complex battery models offline using historical data before deployment. The trained model parameters are then stored and used for online inference, eliminating the need for real-time complex calculations on the embedded controller while maintaining high accuracy.

Inventive Principle:
Principle #10Preliminary action

2Measurement precision

If more historic data and referrable data from other battery systems are stored, then measurement precision is improved, but data storage capacity requirements increase

Engineering Contradiction:
ImproveSOC estimation accuracyVSAvoiddata storage capacity
Core Design Contradiction:
Measurement precisionVSQuantity of substance

Solution Approach 1:

The system extracts the data storage function from the embedded controller and places it on an external server. The embedded controller only stores minimal local data, while historical data and data from other battery systems are stored and managed on the server, enabling access to large datasets without increasing on-board storage requirements.

Inventive Principle:
Principle #2Taking out (Extraction)

3Measurement precision

If complex battery models are established and validated using offline testing data, then measurement precision is improved, but loss of time in data processing and model validation increases

Engineering Contradiction:
Improvebattery model accuracyVSAvoidmodel validation time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The system performs model training and validation in advance during an offline phase using historical data. Once the complex model is trained and validated, it is converted to a simplified model format that can be quickly deployed for real-time SOC estimation, eliminating the need for time-consuming validation during online operation.

Inventive Principle:
Principle #10Preliminary action

4Measurement precision

If powerful IC chips are used for real-time calculation of complex battery models, then measurement precision is improved, but device complexity and cost increase

Engineering Contradiction:
ImproveSOC estimation accuracyVSAvoidchip complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The system creates a simplified copy of the complex battery model that can be executed on standard embedded controllers. The complex model's knowledge is transferred to a simpler model structure through offline training, allowing accurate SOC estimation without requiring powerful IC chips in the embedded system.

Inventive Principle:
Principle #26Copying

Data Source

PatentEP3224632B1Wireless network based battery management system
Publication Date: 2019.04.24 ROBERT BOSCH GMBH
  • EP3224632B1 patent drawingFigure 1
  • EP3224632B1 patent drawingFigure 2~4
  • EP3224632B1 patent drawingFigure 5

AI summary

A wireless network based battery management system is provided, which comprises an off-board subsystem and an on-board subsystem, wherein the off-board subsystem comprising an off-board data storage for storing historic data of a on-board battery system and data of other battery systems, and an off-board data processing device for analyzing and processing the stored data and establishing and validating off-board battery models, mapping the accurate and complex off-board battery models into a simple off-board battery model, and generating parameters of the simple off-board battery mode, and wherein the on-board subsystem selects a simple on-board battery mode corresponding to the simple off-board battery mode, updating parameter of the simple on-board battery mode with those of the simple off-board battery mode, and calculates battery status of the on-board battery system. A method of using the wireless network based battery management system described above is also provided.