Safe Online Control-Informed Learning
This addresses safety and adaptation challenges in autonomous systems, representing an incremental improvement by integrating existing methods like extended Kalman filters and barrier functions into a unified framework.
The paper tackles the problem of safety-critical autonomous systems by proposing a Safe Online Control-Informed Learning framework that unifies optimal control, parameter estimation, and safety constraints, achieving robust adaptation with theoretical convergence and safety guarantees demonstrated on cart-pole and robot-arm systems.
This paper proposes a Safe Online Control-Informed Learning framework for safety-critical autonomous systems. The framework unifies optimal control, parameter estimation, and safety constraints into an online learning process. It employs an extended Kalman filter to incrementally update system parameters in real time, enabling robust and data-efficient adaptation under uncertainty. A softplus barrier function enforces constraint satisfaction during learning and control while eliminating the dependence on high-quality initial guesses. Theoretical analysis establishes convergence and safety guarantees, and the framework's effectiveness is demonstrated on cart-pole and robot-arm systems.