Safe and Stable RL (S<sup>2</sup>RL) Driving Policies Using Control Barrier and Control Lyapunov Functions

Gangopadhyay, Briti; Dasgupta, Pallab; Dey, Soumyajit

doi:10.1109/tiv.2022.3160202

Cited by 12 publications

(2 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The backstepping-based BLF [3][4][5]30] approach for nonlinear systems has difficulty in handling dynamic constraints and has more stringent constraints on the initial value of the system state. The previous work on NSDF [7] is also only capable of addressing numerical constraints, and the NSDF desigred in [10,11] leads to complex derivations of the new states.…”

Section: Ensdfmentioning

confidence: 99%

“…In safety-critical systems, exceeding specific state thresholds poses significant risks. Presently, the Barrier Lyapunov Function (BLF) is extensively employed for nonlinear systems with state constraints [3][4][5]. The core concept of the BLF is to directly constrain the virtual control law during the backstepping design process, thereby indirectly ensuring that the state variables do not violate the constraint conditions.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Constraint-Depentdent Switching Event-Triggered Control for UncertainNonlinear Systems with State Constraints

Liu,

Duan,

et al. 2024

Preprint

View full text Add to dashboard Cite

This paper investigates the event-triggered control problem for a class of strict-feedback uncertain nonlinear systems with switching state constraints. The main challenges of such work arise from two aspects: (1) The state is not constrained at all moments, so the constraints on the state are not fixed, resulting in switches between bounded and unbounded constraints. The existing research is more suitable for bounded constraints. (2) Due to the presence of switching constraints, we aim to address these scenarios with varying control performance, existing research addresses this solely via offline tuning of the controller and its parameters. To solve the above problems, we develop a new exponential nonlinear state-dependent function (ENSDF). The ENSDF transforms the constraint problem of the original state into the boundedness problem of the new state, independent of the boundedness of the constraint boundaries. Then, a new constraint-dependent switching event-triggered controller is developed. The switching event-triggered controller based on ENSDF can dynamically adjust the controller and its parameters online to adapt to switching constraints. It is theoretically shown that the state does not violate the constraints, all closed-loop signals are globally bounded, and the tracking error converges towards an adjustable compact set.

show abstract

Section: Ensdfmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Constraint-Depentdent Switching Event-Triggered Control for UncertainNonlinear Systems with State Constraints

Liu,

Duan,

et al. 2024

Preprint

View full text Add to dashboard Cite

show abstract

Navigation for autonomous vehicles via fast-stable and smooth reinforcement learning

Zhang,

Yang,

Liang

et al. 2023

Sci. China Technol. Sci.

View full text Add to dashboard Cite

Recent advances in reinforcement learning-based autonomous driving behavior planning: A survey

Wu,

Huang,

Huang

et al. 2024

Transportation Research Part C: Emerging Technologies

View full text Add to dashboard Cite

Safe and Stable RL (S²RL) Driving Policies Using Control Barrier and Control Lyapunov Functions

Cited by 12 publications

References 26 publications

Constraint-Depentdent Switching Event-Triggered Control for UncertainNonlinear Systems with State Constraints

Constraint-Depentdent Switching Event-Triggered Control for UncertainNonlinear Systems with State Constraints

Navigation for autonomous vehicles via fast-stable and smooth reinforcement learning

Recent advances in reinforcement learning-based autonomous driving behavior planning: A survey

Contact Info

Product

Resources

About

Safe and Stable RL (S2RL) Driving Policies Using Control Barrier and Control Lyapunov Functions

Cited by 12 publications

References 26 publications

Constraint-Depentdent Switching Event-Triggered Control for UncertainNonlinear Systems with State Constraints

Constraint-Depentdent Switching Event-Triggered Control for UncertainNonlinear Systems with State Constraints

Navigation for autonomous vehicles via fast-stable and smooth reinforcement learning

Recent advances in reinforcement learning-based autonomous driving behavior planning: A survey

Contact Info

Product

Resources

About

Safe and Stable RL (S²RL) Driving Policies Using Control Barrier and Control Lyapunov Functions