FinsSim: A Reality-Aligned Integrated Simulation Platform for Underwater Robot Learning
A Sim-to-Real platform connecting high-fidelity simulation, learning workflows, and hardware deployment
Underwater reinforcement learning research, for the motion control of vehicles in complex aquatic environments.
Note
“Chase the sea, and become the sea.”
Research threads
- Simulation-to-Hardware platforms construction
- Pool-world global positioning system
- RL, IRL topics research
- Multiagent collaboration and control
Working modes
- AI Agent for system designing
- ROS 2 and real world system interaction
- Robotics deployment
Active projects
Current projects and experimental threads.
MARL collaboration
Multiagent collaboration project
IRL for underwater control
DPPO and IRL
Related publications
Connected publication frames.
These publications are the public-facing slices that already emerged from this research line.