1. Current Research

1.1. Learning-Based Predictive Control for Active Cell Balancing in Electrified Vehicles for Range Extension

1.1.1. Team Members

Dr. Ali Arshad Uppal (PI, Pakistan), Dr. Qadeer Ahmed (PI, USA), Mr. Afaq Ahmed, Dr. Muhammad Rizwan Azam and Dr. Syed Bilal Javed

1.1.2. Background

The electric vehicle (EV) market is steadily growing, with projections indicating over 672 million EVs on the roads by 2050. The battery pack, primarily lithium-ion, serves as the core of an EV, influencing its range. Cell imbalance, a key factor limiting EV range, can lead to battery pack shutdown based on a single cell’s condition. Active cell balancing (ACB) emerges as a crucial solution to counter cell imbalance, enhancing overall battery pack performance, safety, and consequently, EV range.

1.1.3. Proposed Research

EV battery packs are equipped with a battery management system (BMS), responsible for monitoring and control. The cell equalization circuit, a pivotal BMS component, plays a vital role. This research aims to extend EV range through state-of-charge (SoC)-based ACB, with objectives including:

1. Selecting an appropriate ACB network (ACBN)
2. Developing a mathematical model for a cell, considering SoC, thermal dynamics, and terminal voltage
3. Detailed mathematical modeling of ACBN, encompassing cell dynamics, balancing currents, and power losses
4. Designing a range-maximizing, robust nonlinear model predictive control (NMPC) using the ACBN model
5. Conducting a simulation study to showcase NMPC performance and EV range extension
6. Developing a prototype for real-time implementation. The choice of the lithium-ion cell’s mathematical model is critical. Empirical models are simple but lack electrochemical insight, while physics-informed models are accurate but computationally complex. Utilizing specially designed excitation signals, experimental data can be employed to develop a data-driven model for the ACBN, achieving a compromise between accuracy and computational cost.

1.1.4. Intellectual Merit

Despite extensive research in ACB control, certain avenues require exploration. Research gaps include:

• Often overlooked high-fidelity modeling of balancing currents and power losses, crucial for assessing ACBN efficiency
• The subjective choice of a cell dynamics model, necessitating alignment with both ACB (local) and EV range extension (global) paradigms
• The nascent stage of research on range extension-aware NMPC, demanding further development.

1.1.5. Broader Impact

The control system’s design for EV battery packs extends beyond immediate applications, offering broader impacts such as:

• Advancing scientific knowledge by combining data-driven and model-based methods to understand and control battery systems, capturing their complex dynamics.
• Enhancing EV industry innovation and competitiveness through a novel technique optimizing battery performance and safety under diverse conditions.
• Improving environmental and social benefits of EVs by reducing emissions, energy consumption, and battery degradation, while increasing driving range, reliability, and affordability

1.1.6. Research Contributions

• Model-Based Quantitative Analysis of a Capacitive Cell Balancing Technique using SoC Estimator
• Power Losses Aware Nonlinear Model Predictive Control Design for Active Cell Balancing

2. Completed Projects

2.1. Machine Learning Based Control-Oriented Modeling of Underground Coal Gaisfication (UCG) Process

2.1.1 Team Members

Dr. Ali Arshad Uppal (PI), Mr. Afaq Ahmed and Dr. Syed Bilal Javed

2.1.2 Research Contributions

• Nonlinear-Control-Oriented Modeling of the MultiVariable Underground Coal Gasification Process for UCG Project Thar: A Machine Learning Perspective
• Development of CAVLAB–A Control-Oriented MATLAB Based Simulator for an Underground Coal Gasification Process
• CAVLAB simulator