A hybrid time-frequency domain system for accurate estimation and thermal management of battery cell state

Background

Lithium-ion batteries have become the cornerstone of modern electric vehicles (EVs) due to their high energy density, long cycle life, and relatively low self-discharge rates. As EV adoption accelerates, the accurate monitoring and management of battery health and performance have become increasingly important. Key to this is the precise estimation of the battery’s state of charge (SOC) and thermal characteristics, which are critical for ensuring safety, optimizing performance, and prolonging battery lifespan. Effective thermal management is essential, as excessive heat can accelerate battery degradation and even lead to hazardous conditions such as thermal runaway. Therefore, advanced diagnostic tools that provide real-time, granular insights into battery state are in high demand within the automotive and energy storage industries.

Current approaches to SOC and thermal state estimation often rely on indirect measurements, such as voltage and current monitoring, or on slow equilibrium-based methods that require the battery to rest for extended periods. These techniques are susceptible to errors caused by voltage drift, overpotential, and the influence of irreversible heat generation, which can obscure the true reversible heat characteristics of the cell. Furthermore, traditional resistance-based thermal models fail to capture the nuanced, material-level behaviors of battery electrodes, limiting the accuracy of thermal predictions and SOC estimates. As a result, these conventional methods can lead to suboptimal battery management, reduced safety margins, and diminished overall performance, highlighting the urgent need for more sophisticated and responsive diagnostic solutions.

Technology

This technology is an advanced battery state estimation system designed for lithium-ion cells, particularly in electric vehicles. It integrates temperature sensors and Peltier elements with each battery cell to apply controlled thermal perturbations—such as sinusoidal, triangular, or square waves—typically with a 5°C amplitude during low load conditions. The system initially measures the cell’s temperature to correct for voltage drift, then applies the thermal perturbation and records the resulting temperature and voltage responses over a set period. These time-domain signals are converted into the frequency domain, often using Fourier transforms, to calculate the entropy coefficient—a key indicator of reversible heat generation. Simultaneously, the system identifies voltage plateau regions that correspond to distinct state of charge (SOC) levels. By combining these analyses, the system generates refined SOC estimates and further decomposes them into material-level parameters, such as the lithiation states and entropy coefficients of the anode and cathode, enabling detailed insights into battery behavior.

What differentiates this technology is its hybrid use of time and frequency domain analysis to deliver rapid and highly accurate SOC and thermal state estimations, overcoming the limitations of traditional methods that are slow and susceptible to voltage drift and overpotential effects. The application of controlled thermal perturbations via Peltier elements, combined with frequency-domain signal processing, allows the system to isolate reversible heat effects not captured by resistance-based models. This enables precise prediction of reversible heat generation, detection of coolant failures, and determination of safe power limits, directly informing real-time battery management and adaptive driving strategies. Its adaptability to various cell chemistries, configurations, and perturbation waveforms, along with its capacity for deployment both onboard and via remote servers, makes it a robust and versatile solution for enhancing electric vehicle safety, efficiency, and battery longevity.

Benefits

• Accurate and fast estimation of lithium-ion battery state of charge (SOC) using thermal perturbations and frequency domain analysis.
• Improved thermal management through precise prediction of reversible heat generation at the material (anode/cathode) level.
• Mitigation of voltage drift and overpotential effects for more reliable battery state measurements.
• Detection of coolant system failures and determination of safe maximum power draw to prevent thermal degradation or runaway.
• Adaptability to various battery chemistries, cell configurations, and perturbation waveforms for broad applicability.
• Enhanced battery performance and longevity by enabling real-time, material-level monitoring and control.
• Support for adaptive driving strategies in fault or limp-home scenarios based on entropy coefficient data.

Applications

• Electric vehicle battery health monitoring
• Thermal management system optimization
• Coolant failure detection
• Adaptive power draw control
• Material-level battery diagnostics

Additional info

This system estimates battery state and thermal properties using controlled thermal perturbations (e.g., 5°C sinusoidal waves) applied by Peltier elements. It measures temperature and voltage responses, transforming them to the frequency domain via Fourier transform to calculate entropy coefficients. Voltage plateaus refine SOC. Material-level parameters (anode/cathode lithiation) are derived for precise thermal management and reversible heat prediction.