Show simple item record

dc.contributor.authorMysore, Siddharthen_US
dc.contributor.authorMabsout, Basselen_US
dc.contributor.authorMancuso, Renatoen_US
dc.contributor.authorSaenko, Kateen_US
dc.date.accessioned2021-09-17T13:40:30Z
dc.date.available2021-09-17T13:40:30Z
dc.date.issued2020
dc.identifier.citationSiddharth Mysore, Bassel Mabsout, Renato Mancuso, Kate Saenko. 2020. "Regularizing Action Policies for Smooth Control with Reinforcement Learning.." CoRR, Volume abs/2012.06644, https://arxiv.org/abs/2012.06644
dc.identifier.urihttps://hdl.handle.net/2144/43029
dc.description.abstractA critical problem with the practical utility of controllers trained with deep Reinforcement Learning (RL) is the notable lack of smoothness in the actions learned by the RL policies. This trend often presents itself in the form of control signal oscillation and can result in poor control, high power consumption, and undue system wear. We introduce Conditioning for Action Policy Smoothness (CAPS), an effective yet intuitive regularization on action policies, which offers consistent improvement in the smoothness of the learned state-toaction mappings of neural network controllers, reflected in the elimination of high-frequency components in the control signal. Tested on a real system, improvements in controller smoothness on a quadrotor drone resulted in an almost 80% reduction in power consumption while consistently training flight-worthy controllers.en_US
dc.language.isoen_US
dc.relation.ispartofCoRR
dc.titleRegularizing action policies for smooth control with reinforcement learningen_US
dc.typeArticleen_US
dc.description.versionFirst author draften_US
pubs.elements-sourcedblpen_US
pubs.notesEmbargo: No embargoen_US
pubs.organisational-groupBoston Universityen_US
pubs.organisational-groupBoston University, College of Arts & Sciencesen_US
pubs.organisational-groupBoston University, College of Arts & Sciences, Department of Computer Scienceen_US
dc.identifier.mycv576678


This item appears in the following Collection(s)

Show simple item record