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

Learning has propelled the cutting edge of performance in robotic control tonew heights, allowing robots to operate with high performance in conditionsthat were previously unimaginable. The majority of the work, however, assumesthat the unknown parts are static or slowly changing. This limits them tostatic or slowly changing environments. However, in the real world, a robot mayexperience various unknown conditions. This paper presents a method to extendan existing single mode GP-based safe learning controller to learn anincreasing number of non-linear models for the robot dynamics. We show thatthis approach enables a robot to re-use past experience from a large number ofpreviously visited operating conditions, and to safely adapt when a new anddistinct operating condition is encountered. This allows the robot to achievesafety and high performance in an large number of operating conditions that donot have to be specified ahead of time. Our approach runs independently fromthe controller, imposing no additional computation time on the control loopregardless of the number of previous operating conditions considered. Wedemonstrate the effectiveness of our approach in experiment on a 900\,kg groundrobot with both physical and artificial changes to its dynamics. All of ourexperiments are conducted using vision for localization.

Experience Recommendation for Long Term Safe Learning-based Model Predictive Control in Changing Operating Conditions