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

Precise trajectory tracking for legged robots can be challenging due to theirhigh degrees of freedom, unmodeled nonlinear dynamics, or random disturbancesfrom the environment. A commonly adopted solution to overcome these challengesis to use optimization-based algorithms and approximate the system with asimplified, reduced-order model. Additionally, deep neural networks arebecoming a more promising option for achieving agile and robust leggedlocomotion. These approaches, however, either require large amounts of onboardcalculations or the collection of millions of data points from a single robot.To address these problems and improve tracking performance, this paper proposesa method based on iterative learning control. This method lets a robot learnfrom its own mistakes by exploiting the repetitive nature of legged locomotionwithin only a few trials. Then, a torque library is created as a lookup tableso that the robot does not need to repeat calculations or learn the same skillover and over again. This process resembles how animals learn their musclememories in nature. The proposed method is tested on the A1 robot in asimulated environment, and it allows the robot to pronk at different speedswhile precisely following the reference trajectories without heavycalculations.

Practice Makes Perfect: an iterative approach to achieve precise tracking for legged robots