▲Wirig Diagram Hot
▲ ▲ ▼

Enhancing EV Range Prediction: Unscented Kalman Filter for Battery State of Charge Estimation

By

Electric Vehicle (EV) adoption is soaring worldwide, driven by the need for sustainable mobility. A critical aspect of EV operation and management is accurately estimating the State of Charge (SoC) of its batteries, crucial for range estimation, optimal charging, and battery longevity. Traditional SoC estimation methods often fall short, leading to the adoption of advanced filtering techniques like the UnscentedKalman Filter (UKF).

UKF offers a powerful alternative to the Extended Kalman Filter (EKF), especially for nonlinear systems like EVs. Unlike EKF, UKF doesn't rely on linearization and thus provides more accurate estimates. It uses a set of carefully chosen sample points (Sigma Points) to capture the statistical properties of the random variables, making it robust and efficient.

Unscented Kalman Filter Basics

The Unscented Kalman Filter operates on the principle of propagated uncertainty. It uses a deterministic sampling technique to approximate the Gaussian random variable's statistics. The algorithm is initialized with an a priori estimate and then iteratively updated with new measurements.

Domain knowledge-guided machine learning framework for state of health estimation in Lithium-ion batteries - Communications Engineering
Domain knowledge-guided machine learning framework for state of health estimation in Lithium-ion batteries - Communications Engineering
Battery Voltage State of Charge
Battery Voltage State of Charge
Electric Vehicle Charging Time Calculator | Accurate EV Charging Time Estimation Tool for Fast Pl...
Electric Vehicle Charging Time Calculator | Accurate EV Charging Time Estimation Tool for Fast Pl...

Despite its superior performance, UKF hasn't been extensively used in real-world EV applications due to the complexity of battery models and the nonlinear nature of battery dynamics. However, recent advancements in battery research and computational power have paved the way for its integration.

Modeling Battery Dynamics with UKF

Battery dynamics can be modeled using electrochemical models or equivalent circuit models. The UKF can handle these complex models by treating the states in the model as random variables. It estimates both the state variables (like SoC,ințăcurrent) and the model parameters (like resistance, capacitance).

the diagram shows how an electric generator works and what it uses to work with other devices
the diagram shows how an electric generator works and what it uses to work with other devices
the diagram shows how lithium batteries work
the diagram shows how lithium batteries work
Lithium-Ion Batteries in Embedded Systems
Lithium-Ion Batteries in Embedded Systems

For instance, in an equivalent circuit model, the UKF can estimate the SoC, terminal voltage, current, and internal resistance. This allows for real-time monitoring of battery health and predictive maintenance.

UKF for SoC Estimation

SoC is typically estimated using the Coulomb counting method, which integrates the current over time. However, this method suffers from cumulative errors due to measurement inaccuracies and varying current rates. UKF can mitigate these errors by fusing SoC estimates from Coulomb counting with other sensor data like terminal voltage and temperature.

Moreover, UKF can account for sensor fusion even in scenarios with missing data or outliers. By incorporating a process noise that models the uncertainty in the SoC estimation, UKF can provide reliable estimates even in challenging operating conditions.

the phase diagram for voltages and current currents is shown in figure 1, 2
the phase diagram for voltages and current currents is shown in figure 1, 2
A hierarchical fusion framework for vehicle to grid energy management using predictive intelligence and learning based pricing - Scientific Reports
A hierarchical fusion framework for vehicle to grid energy management using predictive intelligence and learning based pricing - Scientific Reports
Estimation of state of health for lithium-ion batteries using advanced data-driven techniques - Scientific Reports
Estimation of state of health for lithium-ion batteries using advanced data-driven techniques - Scientific Reports
three green batteries sitting on top of a piece of electronic circuit board covered in tiny dots
three green batteries sitting on top of a piece of electronic circuit board covered in tiny dots
Warning EV Battery Degradation Signs Every U.S. Driver Must Know
Warning EV Battery Degradation Signs Every U.S. Driver Must Know
How close are we to solid state batteries for electric vehicles?
How close are we to solid state batteries for electric vehicles?

Challenges and Future Directions

While UKF shows great promise, there are still several challenges to be addressed before it can be widely adopted in EVs. One key challenge is the representativeness of the.System model. The performance of UKF is heavily dependent on the model accuracy, and inaccuracies can lead to drifted estimates.

Furthermore, the computational complexity of UKF can be high, limiting its real-time implementation in resource-constrained embedded systems. Real-time implementation of UKF in EVs requires efficient algorithms and hardware architecture optimization.

Despite these challenges, the future looks bright for UKF-based SoC estimation in EVs. With advancements in battery research, the computational power of onboard controllers is increasing, enabling more complex algorithms. Moreover, machine learning techniques can help develop more accurate battery models, further improving UKF's performance.

LiTime 80 Amp Lithium Charger - Big Power & Fast Charging - I Love RV Life
LiTime 80 Amp Lithium Charger - Big Power & Fast Charging - I Love RV Life
💬 Comments ↗ Share ★ Save
AD