Battery Life Prediction Model Using Time-Series and Discrete Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods lack an effective and efficient way to predict the life of secondary batteries using artificial intelligence models, which is crucial for optimizing battery performance and maintenance.
Innovation Solution
A method and apparatus using a battery life prediction model, an artificial intelligence model, generate combined vectors from time series and discrete data related to charging and discharging, and determine a life prediction value based on capacity prediction values.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional methods are used to predict battery life, then the process is simple, but the prediction accuracy is insufficient
Solution Approach 1:
The patent segments battery data into two distinct types: time series data (capturing temporal patterns in voltage, current, temperature) and discrete data (capturing categorical features like charge/discharge cycles, battery type). This segmentation allows each data type to be processed by specialized subvectors within the embedding layer, improving prediction accuracy while maintaining manageable model complexity through targeted feature extraction.
Solution Approach 2:
The patent introduces an embedding layer as an intermediary component that transforms both time series data and discrete data into a unified vector representation. This embedding layer acts as a mediator that reconciles the different data types and their characteristics, enabling the neural network to process heterogeneous battery data effectively without overwhelming complexity.
2Measurement precision
If comprehensive data processing is performed to improve prediction accuracy, then the model becomes more accurate, but the processing time increases
Solution Approach 1:
The patent performs preliminary data processing by creating embeddings for both time series data and discrete data before feeding them into the neural network. The embedding layer pre-processes and transforms raw data into meaningful vector representations, capturing essential patterns and relationships in advance. This preliminary action reduces the computational burden during actual prediction, balancing accuracy with processing efficiency.
3Adaptability or versatility
If separate processing is used for different data types, then the model can handle diverse data, but the overall system complexity increases
Solution Approach 1:
The patent merges the processing of time series data and discrete data into a unified neural network architecture. Both data types are transformed into vector representations and combined in the embedding layer, allowing the model to handle diverse battery data effectively. This merging approach maintains adaptability while avoiding the need for separate complex processing systems for each data type.
Solution Approach 2:
The embedding layer serves as a universal component that handles multiple data types (time series and discrete data) through a single mechanism. This multi-functional approach allows the model to process diverse battery characteristics uniformly, enhancing versatility without proportionally increasing system complexity.
Data Source
AI summary
The apparatus for predicting a life of a battery by using an artificial intelligence model includes a memory storing at least one program and at least one processor configured to execute the at least one program to generate a combined vector by using subvectors derived from time series data and discrete data, which are related to charging and discharging of the battery, input, as input data, the generated combined vector to a battery life prediction model and determine whether a capacity prediction value of the battery, obtained as output data of the battery life prediction model, is less than or equal to a preset value, and derive a life prediction value for the battery in response to a result of determining that the capacity prediction value of the battery is less than or equal to the preset value.