Whole-Body Safe Control of Robotic Systems with Koopman Neural Dynamics
This work addresses the problem of safe and efficient control for robotic systems with complex dynamics, offering a practical solution for real-time applications, though it appears incremental as it combines existing techniques like Koopman operators and SSA.
The paper tackles the challenge of real-time safe control for robots with nonlinear dynamics by proposing a data-driven framework that learns a Koopman embedding and integrates it with the Safe Set Algorithm, enabling efficient linear control via a single quadratic program. It validates the method on a Kinova Gen3 manipulator and a Go2 quadruped, showing accurate tracking and obstacle avoidance.
Controlling robots with strongly nonlinear, high-dimensional dynamics remains challenging, as direct nonlinear optimization with safety constraints is often intractable in real time. The Koopman operator offers a way to represent nonlinear systems linearly in a lifted space, enabling the use of efficient linear control. We propose a data-driven framework that learns a Koopman embedding and operator from data, and integrates the resulting linear model with the Safe Set Algorithm (SSA). This allows the tracking and safety constraints to be solved in a single quadratic program (QP), ensuring feasibility and optimality without a separate safety filter. We validate the method on a Kinova Gen3 manipulator and a Go2 quadruped, showing accurate tracking and obstacle avoidance.