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 Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Model-free or learning-based control, in particular, reinforcement learning(RL), is expected to be applied for complex robotic tasks. Traditional RLrequires a policy to be optimized is state-dependent, that means, the policy isa kind of feedback (FB) controllers. Due to the necessity of correct stateobservation in such a FB controller, it is sensitive to sensing failures. Toalleviate this drawback of the FB controllers, feedback error learningintegrates one of them with a feedforward (FF) controller. RL can be improvedby dealing with the FB/FF policies, but to the best of our knowledge, amethodology for learning them in a unified manner has not been developed. Inthis paper, we propose a new optimization problem for optimizing both the FB/FFpolicies simultaneously. Inspired by control as inference, the optimizationproblem considers minimization/maximization of divergences between trajectory,predicted by the composed policy and a stochastic dynamics model, andoptimal/non-optimal trajectories. By approximating the stochastic dynamicsmodel using variational method, we naturally derive a regularization betweenthe FB/FF policies. In numerical simulations and a robot experiment, weverified that the proposed method can stably optimize the composed policy evenwith the different learning law from the traditional RL. In addition, wedemonstrated that the FF policy is robust to the sensing failures and can holdthe optimal motion. Attached video is also uploaded on youtube:https://youtu.be/zLL4uXIRmrE

Optimization Algorithm for Feedback and Feedforward Policies towards Robot Control Robust to Sensing Failures