Zejian Zhou
University of Wyoming, WY·Engineering
Autonomous systems such as self-driving cars, power grids and robots must keep working when something goes wrong. A disruption can be harmless sensor noise, a broken part, a gust of wind, or a deliberate cyber-attack.
- Fields
Engineering
Computer Science
- Topics
autonomous systems
connected and autonomous vehicles
reinforcement learning
- Grants
- 1 known grant, through Jul 2027
What they work on
- Model complex autonomous systems as multi-agent systems in which a disrupted component becomes an irrational agent, either passive (an accidental fault) or active (adversarial, such as a cyber-attack).
- Develop a cooperative index that quantifies how far a component deviates from its expected cooperative behavior, serving as a real-time signal that a disruption has occurred.
- Use inverse reinforcement learning, which infers an agent's objectives from its observed actions, to distinguish passive faults from active adversaries and improve joint rewards across the system.
- Build an Actor-Critic-Friend-or-Foe (ACFoF) algorithm that dynamically updates control policies during disruptions while maintaining performance, with connected and autonomous vehicles as the application.
More about the research
Today's control algorithms generally cannot tell these causes apart on their own, so they respond poorly or need a human to step in, which leads to degraded performance and possible failures. This project asks how a machine can recognize what kind of disruption it faces and adjust its behavior in real time. The main test case is connected and autonomous vehicles.
Why it matters
More reliable self-correcting control could make connected and autonomous vehicles safer when sensors fail or attackers interfere. The same ideas may carry over to power grids, robotics, intelligent transportation, space exploration and disaster response, where systems must handle unexpected faults or attacks without waiting for a human operator.
Known grants
From NIH and NSF research grants. Labs often have other funding we can't see.
Closest labs
Hao Xu
University of Nevada-Reno, NV·Engineering·Early-career
Develop a hierarchical game theory (HGT) for large-scale multi-agent systems, such as power management and transportation networks, that maintains optimal efficiency while keeping computing demands manageable as agent numbers grow.
Sandhya Saisubramanian
Oregon State University, OR·Computer Science·Early-career
Develop model cognizance methods that let autonomous agents proactively identify and repair limitations in their symbolic decision-making models through limited user interaction.
Junshan Zhang
University of California-Davis, CA·Computer Science
Build CarDreamer, an open-source ecosystem of modular tools for world-model-based self-supervised learning in autonomous driving, through community scoping and planning with academic and industry users.
The school
University of Wyoming: 8 labs on PI Pages, $167M a year on research, 108 research doctorates a year.
Reach out
A start for your email. Fill in the brackets.