Geometric Control of Nonholonomic Agents
Use of Projected Gradients for the Control of Single-agent and Consensus of Multiagent systems.
Guide: Prof. Ravi N. Banavar (IIT Bombay) and Prof. Zhong-ping Jiang (New York University).
- Designed appropriate Morse function to achieve asymptotic convergence to a position for a single agent
- Utilized switching control to introduce time-varying and discontinuous feedback to achieve exponential convergence to a position for a nonholonomic agent in simulations
- Extended the controller for a multi-agent system, achieving exponential position-consensus
System and controller visualization