Pith. sign in

REVIEW

Suboptimal Safety-Critical Control for Continuous Systems Using Prediction-Correction Online Optimization

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2203.15305 v3 pith:DG6LPL6J submitted 2022-03-29 eess.SY cs.SY

Suboptimal Safety-Critical Control for Continuous Systems Using Prediction-Correction Online Optimization

classification eess.SY cs.SY
keywords algorithmscontrolsuboptimalcontinuousmethodoptimizationsolutionsconvergence
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

This paper investigates the control barrier function (CBF) based safety-critical control for continuous nonlinear control affine systems using the more efficient online algorithms through time-varying optimization. The idea lies in that when quadratic programming (QP) or other convex optimization algorithms needed in the CBF-based method is not computation affordable, the alternative suboptimal feasible solutions can be obtained more economically. By using the barrier-based interior point method, the constrained CBF-QP problems are converted into the unconstrained ones with suboptimal solutions tracked by two continuous descent-based algorithms. Considering the lag effect of tracking and exploiting the system information, the prediction method is added to the algorithms which thereby achieves a exponential convergence rate to the time-varying suboptimal solutions. The convergence and robustness of the designed methods as well as the safety criteria of the algorithms are analyzed theoretically. In the end, the effectiveness is illustrated by simulations on the anti-swing and obstacle avoidance tasks.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.