EV Battery Voltage Compression and Validation Using AI Encoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
As high-voltage battery packs in electric vehicles increase in size, the volume of measurement data for battery cell voltages becomes significant, consuming controller and communication resources and vulnerable to hacking, with existing technologies failing to efficiently compress, encrypt, and validate this data effectively.
Innovation Solution
Implementing a neural network-based system that compresses, encrypts, and decompresses battery cell voltage measurements within the vehicle, using a symmetric neural network or autoencoder to generate a compressed encoded representation, which can be reconstructed for control purposes, and includes a diagnostic feature to validate data integrity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If all battery cell voltage measurements are transmitted to controllers, then data validity can be verified, but controller and communication network resources are excessively consumed
Solution Approach 1:
The patent extracts only the essential information from full battery cell voltage measurements by generating hash values. Instead of transmitting complete voltage data sets, the system computes and transmits only hash representations, which are sufficient for validity verification but consume minimal communication and processing resources.
Solution Approach 2:
The patent transforms the data representation from complete voltage measurements to hash values through a mathematical transformation process. This parameter change reduces the data size and complexity while preserving the essential verification capability, allowing controllers to validate data without handling the full data set.
2Ease of operation
If battery cell voltage data is transmitted over communication networks, then vehicle controllers can access the data, but the data becomes vulnerable to snooping and hacking
Solution Approach 1:
The patent introduces hash values as an intermediary between the original battery cell voltage data and the vehicle controllers. This intermediary layer allows controllers to verify data validity without directly accessing the sensitive raw voltage measurements, thereby maintaining data security while enabling necessary verification operations.
Solution Approach 2:
The patent creates a cryptographic copy (hash value) of the original battery cell voltage data that preserves verification functionality without exposing the original sensitive information. This copy can be freely transmitted and verified, while the original data remains protected and inaccessible to unauthorized parties.
3Loss of energy
If compressed encoded representations are used for data transmission, then communication resources are conserved, but data integrity validation becomes more complex
Solution Approach 1:
The patent implements a self-service validation mechanism where the hash value inherently contains the verification capability. The compressed encoded representation is designed so that its own structure enables integrity validation without requiring external complex verification processes, thus conserving communication resources while maintaining simple validation.
Data Source
AI summary
A vehicle includes a traction battery and a cell monitor associated with battery cell strings each having associated connected battery cells. Each monitor is configured to generate a compressed encoded representation of the traction battery cell voltage measurements using artificial intelligence, such as a neural network. A battery controller in communication with each cell monitor receives the compressed encoded representation and generates reconstructed battery cell voltage measurements and controls the traction battery in response to the reconstructed traction battery cell voltage measurements. An unencoded battery cell voltage measurement may be communicated with the compressed encoded representation of the battery cell voltage measurements and used to validate the reconstructed measurements.