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

Although deep reinforcement learning (deep RL) methods have lots of strengthsthat are favorable if applied to autonomous driving, real deep RL applicationsin autonomous driving have been slowed down by the modeling gap between thesource (training) domain and the target (deployment) domain. Unlike currentpolicy transfer approaches, which generally limit to the usage ofuninterpretable neural network representations as the transferred features, wepropose to transfer concrete kinematic quantities in autonomous driving. Theproposed robust-control-based (RC) generic transfer architecture, which we callRL-RC, incorporates a transferable hierarchical RL trajectory planner and arobust tracking controller based on disturbance observer (DOB). The deep RLpolicies trained with known nominal dynamics model are transfered directly tothe target domain, DOB-based robust tracking control is applied to tackle themodeling gap including the vehicle dynamics errors and the externaldisturbances such as side forces. We provide simulations validating thecapability of the proposed method to achieve zero-shot transfer across multipledriving scenarios such as lane keeping, lane changing and obstacle avoidance.

Zero-shot Deep Reinforcement Learning Driving Policy Transfer for Autonomous Vehicles based on Robust Control