Lithium Battery SOC Estimation via Clustering Sub-Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for estimating the state of charge (SOC) of lithium batteries are inaccurate and unreliable due to the use of single global models that fail to represent local process characteristics under various working conditions.
Innovation Solution
A method that involves collecting state data and corresponding SOC values under different conditions, performing clustering analysis to create sample subsets, establishing sub-models for each subset, selecting the most relevant sub-model based on change values, and calculating the SOC value using weighted sub-model functions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If a single global model is used for SOC estimation, then the model structure is simple and easy to implement, but the accuracy and reliability of SOC estimation deteriorates because it cannot represent local process characteristics under multiple working conditions
Solution Approach 1:
The patent divides the sample set into multiple sample subsets based on different working conditions (charging/discharging rates, temperature ranges). Each subset is used to train a separate sub-model, allowing the system to capture local characteristics of different operating regimes rather than averaging them in a single global model.
Solution Approach 2:
Different sub-models are trained on different working condition subsets, enabling each model to specialize in local characteristics of specific operating ranges. The system then selects the most appropriate sub-model based on current working conditions, ensuring locally optimized estimation accuracy rather than a compromise global performance.
2Measurement precision
If multiple sub-models are established for different sample subsets, then the SOC estimation accuracy improves by representing local characteristics, but the device complexity increases
Solution Approach 1:
The system dynamically selects which sub-model to use based on real-time working conditions (charging/discharging current, temperature). This dynamic adaptation allows the system to maintain high accuracy across varying conditions without permanently maintaining multiple active models, reducing the effective complexity at any given moment.
Solution Approach 2:
The patent introduces a working condition judgment module as an intermediary that determines which sub-model should be applied based on current operating parameters. This mediator simplifies the complexity by providing a clear decision-making layer that maps working conditions to appropriate models, rather than requiring direct complex interactions between multiple models.
3Reliability
If multiple sub-models are used with weighted calculations, then the SOC estimation reliability improves, but the calculation complexity increases
Solution Approach 1:
Instead of always using all available sub-models with equal weight, the system applies only the most relevant sub-model(s) based on current working conditions. This partial action approach maintains reliability by using the most appropriate model while reducing calculation complexity by excluding irrelevant models from the computation.
Data Source
Figure 1
Figure 2
Figure 3~4
AI summary
The present application relates to the technical field of SOC estimation of a lithium battery, and discloses a method, device and computer readable storage medium for estimating SOC of a lithium battery. State data and corresponding SOC values of lithium batteries under different working conditions are collected to establish a sample set, and clustering analysis is performed on the sample set to obtain a plurality of sample subsets; then a corresponding sub-model is established for each of the sample subsets to obtain sub-model functions of the plurality of sample subsets; next, the state data of a sample to be tested is respectively added into the state data of each of the sample subsets to calculate a change value of the state data of each of the sample subsets before and after the adding operation, and at least one sub-model close to the sample to be tested is selected as the selected sub-model according to the change value; and finally, a weight is assigned to the selected sub-model to calculate the SOC value of the sample to be tested. By obtaining the estimated SOC value of the lithium battery in the aforementioned manner, the accuracy and reliability of the estimated SOC value of the lithium battery can be improved.