Applications
Cell selection
Compare cells in the context of your application. Use our list of metrics: SOC range, cell voltage, heat generation, energy throughput or create your own.
Pack architecture
Speed up design iterations and lower costs. Add cells in series or parallel, connect auxiliary systems, and try out multiple drive cycles or operating conditions.
Thermal performance
Use heat generation and temperature dependance to design and improve thermal management systems. Use the provided lumped thermal model, or couple to your existing thermal models.
BMS
Combine our models & parameter sets with state estimation algorithms such as Kalman filters for SoC/SoH estimation. Design and evaluate BMS systems or just upload our ECM tables to off-the-shelf solutions.
Key features
Full validation
Our validation experimental data is never used to train our models. This prevents over-fitting and ensures robustness across operating conditions. Our validation cycles are typically adapted from real-world drive cycles and mission profiles.
Leading accuracy
Our ECMs predict the terminal voltage to impressive levels of accuracy over a wide operating range. We take extra care to fit globally using a robust process, leading to results you can trust.
Industry-standard 2-RC pair structure
We use Resistor and Capacitor (RC) pairs in our ECMs, capturing cell losses and dynamic response to pulse loads. Model parameters are provided as functions of SoC and temperature. Look-up tables always provided as .csv files too.
Utility features
Electrical models are provided with a range of additional utility features such as adaptive (CV hold) charging and current/power loads.
Fast computation time
We have optimised our Simscape implementation to be extremely computationally efficient.
Easily create a pack
Simply drag and drop multiple blocks into your Simscape environment to model pack behaviour. Integrate other Simscape library components such as resistors, power electronics, and thermal management.
Thermal model
Our ECMs come with a lumped thermal model as standard, with all parameters including temperature-dependence.
Industry leading parameterisation
Our measurements are made using a proprietary sequence of loading conditions using industry leading thermal control and measurement. To maintain high-quality data, we employ proprietary thermal control systems instead of traditional convection-based climatic chambers, ensuring uniform temperature conditions to facilitate the separation of electrical and thermal battery phenomena.
Optional extras
Hysterisis
Including hysteresis in an ECM enhances the accuracy of simulations for charging and discharging cycles. It accounts for the battery's open-circuit voltage response and state of charge history, improving predictions under varied conditions.
Custom Validation
We also work with our customers to tailor our model validation to their specific use cases.
Import into your platform of choice
We can help you import our lookup tables into any platform you choose. Thanks to the standard table format we can integrate into your workflow.
Higher-order thermal models
Couple our ECM to a more advanced thermal models to capture temperature gradients through the cell.
Proprietary cells
We are trusted by some of the world’s biggest OEMs to keep proprietary data safe.
Downloadable as
Look-up table
Parameters only
.png)
Simscape
Object file
What is an ECM?
Equivalent Circuit Models (ECM) predict key cell-level quantities, such as the voltage-current relationship and cell heat generation. They consist a simple electrical circuit which replicate the voltage response of the cell, avoiding the need to model internal cell electrochemical processes.
They are typically very accurate within specified conditions and fast to solve, allowing engineers to evaluate their battery design and performance at a fraction of the cost and time.
Predicts
Voltage
Heat generation
Accounts for
State of Charge (SOC)
Temperature
Accuracy is key
Check out an example validation plot and the extremely high-accuracy of our models. Note that our models are never trained on the validation datasets.
