EV Battery Model Calibration Across SOC and Temperature
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing battery models for electric vehicles face challenges in calibration due to non-smooth parameter distribution, lack of physical principles for temperature dependency, and time-consuming processes, leading to inaccurate and discontinuous performance estimates.
Innovation Solution
A global calibration approach is implemented using a pre-defined parameter-dependency structure, where parameter values are expressed as smoothed functions of state of charge (SOC) and temperature, governed by the Arrhenius equation, to ensure continuous and reliable model performance across SOC and temperature ranges.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If discrete curve fitting or optimization approaches are used to identify parameters at specific SOC or temperature set points, then calibration can be performed with limited data, but non-smooth parameter distribution is induced that creates interpolation errors
Solution Approach 1:
The patent transforms the parameter identification approach by changing from discrete point-wise parameter identification to continuous functional parameter representation. Parameters are expressed as smooth functions of SOC and temperature using polynomial expansions, which inherently provide smooth distributions across the entire operating range while maintaining calibration feasibility with limited data points.
Solution Approach 2:
The patent applies preliminary action by pre-defining the functional form of parameter dependencies on SOC and temperature before calibration. The polynomial expansion structures are established in advance with predetermined orders, which guides the calibration process to produce smooth parameter distributions without requiring extensive post-processing or interpolation.
2Device complexity
If traditional calibration methods are used without physical principles, then calibration process can be simplified, but temperature dependency lacks physical basis leading to inaccurate models
Solution Approach 1:
The patent introduces physical principle-based intermediary relationships between temperature and model parameters through the Arrhenius equation. This intermediary physical model provides a scientifically grounded connection between temperature variations and parameter changes, enhancing model reliability while maintaining manageable calibration complexity through the use of established physical laws.
3Measurement precision
If comprehensive calibration across all SOC and temperature points is performed, then model accuracy is improved, but calibration time increases significantly
Solution Approach 1:
The patent changes the approach from measuring/calibrating parameters at every discrete SOC and temperature point to using continuous polynomial functions that represent parameters across the entire range. This parameter transformation reduces the number of required calibration measurements from a full grid to a manageable subset, significantly reducing calibration time while maintaining high model accuracy through the smooth functional representations.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach results in improved model reliability and continuity, providing accurate and efficient battery management by smoothing parameter mappings and reducing calibration time.
Implementation Method 1
the temperature effects are governed by the Arrhenius equation
Data Source
AI summary
The present solution provides calibration for a battery model of an electric vehicle battery. One or more processors of an electric vehicle coupled with memory can identify, for the electric vehicle, a model configured to determine performance of a battery of the electric vehicle based on a function. The function can generate values for a parameter of the model across a range of states of charges and a range of temperatures. The one or more processors can generate, based on input from a sensor of the electric vehicle, a value of performance of the electric vehicle via the model. The one or more processors can provide an action for the electric vehicle based on the value of the performance.