A General Safety Framework for Learning-Based Control in Uncertain Robotic Systems

Fisac, Jaime F.; Akametalu, Anayo K.; Zeilinger, Melanie N.; Kaynama, Shahab; Gillula, Jeremy H.; Tomlin, Claire J.

doi:10.48550/arxiv.1705.01292

Cited by 7 publications

(15 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The study of safe learning dates back to the beginning of this century [9]. In [10] and [11], Lyapunov-based reinforcement learning is used to allow a learning agent to safely switch between pre-computed baseline controllers.…”

Section: Related Workmentioning

confidence: 99%

“…In several other papers, including [13], [14] and [15], safe exploration methods are utilized to allow the learning modules to achieve a desired balance between ensuring safe operation and exploring new states for improved performance. In [9], a general framework is proposed for ensuring safety of learning-based control strategies for uncertain robotic systems. In this framework, robust reachability guarantees from control theory are combined with Bayesian analysis based on empirical observations.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Provably Robust Learning-Based Approach for High-Accuracy Tracking Control of Lagrangian Systems

Helwa

Heins

Schoellig

2019

IEEE Robot. Autom. Lett.

View full text Add to dashboard Cite

Lagrangian systems represent a wide range of robotic systems, including manipulators, wheeled and legged robots, and quadrotors. Inverse dynamics control and feedforward linearization techniques are typically used to convert the complex nonlinear dynamics of Lagrangian systems to a set of decoupled double integrators, and then a standard, outer-loop controller can be used to calculate the commanded acceleration for the linearized system. However, these methods typically depend on having a very accurate system model, which is often not available in practice. While this challenge has been addressed in the literature using different learning approaches, most of these approaches do not provide safety guarantees in terms of stability of the learning-based control system. In this paper, we provide a novel, learning-based control approach based on Gaussian processes (GPs) that ensures both stability of the closed-loop system and high-accuracy tracking. We use GPs to approximate the error between the commanded acceleration and the actual acceleration of the system, and then use the predicted mean and variance of the GP to calculate an upper bound on the uncertainty of the linearized model. This uncertainty bound is then used in a robust, outer-loop controller to ensure stability of the overall system. Moreover, we show that the tracking error converges to a ball with a radius that can be made arbitrarily small. Furthermore, we verify the effectiveness of our approach via simulations on a 2 degree-offreedom (DOF) planar manipulator and experimentally on a 6 DOF industrial manipulator.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Provably Robust Learning-Based Approach for High-Accuracy Tracking Control of Lagrangian Systems

Helwa

Heins

Schoellig

2019

IEEE Robot. Autom. Lett.

View full text Add to dashboard Cite

show abstract

“…The computations are based on Lyapunov's method and result in two optimization problems: The first optimization problem defines a quadratic approximation of the nonlinearity in the Lyapunov conditions and the second one describes the computation of the safe set and controller. Similar to [2], [3] the framework can be used to augment any desired controller, which is lacking safety guarantees, in particular one which is based on learning.…”

Section: Introductionmentioning

confidence: 99%

“…Extensions of existing RL methods have been developed to enable safe RL with respect to different notions of safety, see [4] for a survey. A detailed literature review regarding RL, focusing on safety with respect to state and input constraints as also considered in this work, can be found in [3]. There are few results for efficient controller tuning from data with respect to best worst-case performance (also worst-case stability under physical constraints) by Bayesian min-max optimization, see e.g.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Scalable synthesis of safety certificates from data with application to learning-based control

Wabersich

Zeilinger

2018

2018 European Control Conference (ECC)

Self Cite

View full text Add to dashboard Cite

The control of complex systems faces a tradeoff between high performance and safety guarantees, which in particular restricts the application of learning-based methods to safety-critical systems. A recently proposed framework to address this issue is the use of a safety controller, which guarantees to keep the system within a safe region of the state space. This paper introduces efficient techniques for the synthesis of a safe set and control law, which offer improved scalability properties by relying on approximations based on convex optimization problems. The first proposed method requires only an approximate linear system model and Lipschitz continuity of the unknown nonlinear dynamics. The second method extends the results by showing how a Gaussian process prior on the unknown system dynamics can be used in order to reduce conservatism of the resulting safe set. We demonstrate the results with numerical examples, including an autonomous convoy of vehicles.

show abstract

Hamilton-Jacobi reachability: A brief overview and recent advances

Bansal

Chen

Herbert

et al. 2017

2017 IEEE 56th Annual Conference on Decision and Control (CDC)

Self Cite

352

258

View full text Add to dashboard Cite

Hamilton-Jacobi (HJ) reachability analysis is an important formal verification method for guaranteeing performance and safety properties of dynamical systems; it has been applied to many small-scale systems in the past decade. Its advantages include compatibility with general nonlinear system dynamics, formal treatment of bounded disturbances, and the availability of well-developed numerical tools. The main challenge is addressing its exponential computational complexity with respect to the number of state variables. In this tutorial, we present an overview of basic HJ reachability theory and provide instructions for using the most recent numerical tools, including an efficient GPU-parallelized implementation of a Level Set Toolbox for computing reachable sets. In addition, we review some of the current work in high-dimensional HJ reachability to show how the dimensionality challenge can be alleviated via various general theoretical and applicationspecific insights.

show abstract

A General Safety Framework for Learning-Based Control in Uncertain Robotic Systems

Cited by 7 publications

References 0 publications

Provably Robust Learning-Based Approach for High-Accuracy Tracking Control of Lagrangian Systems

Provably Robust Learning-Based Approach for High-Accuracy Tracking Control of Lagrangian Systems

Scalable synthesis of safety certificates from data with application to learning-based control

Hamilton-Jacobi reachability: A brief overview and recent advances

Contact Info

Product

Resources

About