English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Reinforcement learning (RL) has had its fair share of success in contact-richmanipulation tasks but it still lags behind in benefiting from advances inrobot control theory such as impedance control and stability guarantees.Recently, the concept of variable impedance control (VIC) was adopted into RLwith encouraging results. However, the more important issue of stabilityremains unaddressed. To clarify the challenge in stable RL, we introduce theterm all-the-time-stability that unambiguously means that every possiblerollout will be stability certified. Our contribution is a model-free RL methodthat not only adopts VIC but also achieves all-the-time-stability. Building ona recently proposed stable VIC controller as the policy parameterization, weintroduce a novel policy search algorithm that is inspired by Cross-EntropyMethod and inherently guarantees stability. Our experimental studies confirmthe feasibility and usefulness of stability guarantee and also features, to thebest of our knowledge, the first successful application of RL withall-the-time-stability on the benchmark problem of peg-in-hole.

Stability-Guaranteed Reinforcement Learning for Contact-rich Manipulation