Session: 09-01-01: Electrochemical Energy Storage and Conversion Systems I

Paper Number: 173583

A Model Predictive Control Approach to Direct Current Fast Charging Control: Limiting Lithium Plating to Balance Charging Speed and Battery Life 

The widespread adoption of electric vehicles (EVs) is closely tied to the availability of fast, reliable charging solutions. DC fast charging, while essential for reducing charging times and improving user convenience, presents challenges to long-term battery health, including the risk of lithium plating, which can degrade battery performance and safety. To address these challenges, advanced control strategies are needed that can intelligently manage the charging process in real time, maximizing charging speed while maintaining battery integrity and health.

In this session we demonstrate the use of model predictive control (MPC) as a robust and flexible framework for optimizing the DC fast charging of lithium-ion batteries. The primary advantage of MPC lies in its ability to anticipate future system behavior by leveraging predictive models, enabling it to compute optimal control actions that balance multiple objectives and constraints, even for unmeasurable states of interest. In the context of EV charging, MPC dynamically adjusts the charging current to minimize total charging time, while strictly enforcing safety limits that prevent harmful side effects such as lithium plating.

The core of the proposed approach is the real-time optimization capability of MPC. At each time step, the controller solves an optimization problem that considers the current battery state, forecasts of future states, and operational constraints. This allows the charging protocol to be continuously adapted to the evolving condition of the battery, rather than relying on static or rule-based charging profiles. By incorporating constraints that capture the onset of lithium plating and other degradation mechanisms, the MPC framework ensures that the charging process always remains within safe operating boundaries.

While the primary focus of this session is control, the effectiveness of MPC is inherently linked to the fidelity of the underlying battery model. To this end, a high-fidelity, physics-based model is employed to provide accurate predictions of internal battery states, including those not directly measurable, such as lithium concentration gradients and overpotentials. This modeling approach enables the MPC controller to make informed decisions about charging currents, particularly under aggressive fast-charging scenarios where the risk of lithium plating is heightened.

This session will detail the design and implementation of the MPC controller, including the formulation of the optimization problem, constraint handling, and integration with real-time state estimation. Simulation studies will be presented to demonstrate how the MPC-based strategy outperforms conventional charging methods, achieving faster charging without exceeding safety limits. The results highlight the ability of MPC to dynamically navigate the trade-off between charging speed and battery health, offering substantial improvements in both areas.

Presenting Author: Jordan Olson MathWorks

Presenting Author Biography: Jordan is an application engineer at MathWorks specializing in artificial intelligence and advanced control design. Jordan supports customers across a wide range of industries, including aerospace, automotive, energy, and robotics. He holds a B.S. and M.S. in Mechanical Engineering, as well as an M.S. in Electrical Engineering, all from The University of Alabama.

Authors:

Jordan Olson MathWorks
Xiangchun Zhang MathWorks