Electrical Engineering · Control · Robotics
Niusha Sedgh
Electrical engineering student interested in control systems, robotics, and autonomous systems with a focus on combining classical control with AI-driven methods for intelligent autonomous platforms.
My work explores areas such as intelligent control, reinforcement learning, autonomous vehicles, multi-agent systems, sensor fusion, and navigation. I use this website to document my research, engineering projects, publications, and ongoing work.
Current Focus
Developing a Hybrid Control Architecture for Vehicle Platooning
My current focus is my bachelor's thesis, which explores a multi-layer control architecture for autonomous vehicle platooning. The main idea is to combine learning-based decision making with model-based control and a digital-twin layer to improve both performance and safety.
01 · Learning Layer
Reinforcement Learning
Exploring reinforcement learning as a high-level decision-making layer for long-term planning and optimization in vehicle platooning.
02 · Safety Layer
Model-Based Control
Studying MPC and related control methods as a lower-level mechanism for maintaining stability, constraint satisfaction, and safe vehicle operation.
03 · Validation Layer
Digital Twin
Investigating how a digital-twin representation can be used to evaluate proposed actions and estimate potentially unsafe situations before commands are applied to the physical system.
04 · Safety Mechanism
Safe Decision Making
Exploring mechanisms that allow the architecture to fall back to model-based control when a learned action is predicted to violate safety constraints or increase collision risk.
Interested in research or collaboration?
Feel free to get in touch for academic discussions and research opportunities.
Get in Touch