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 Tools annual

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 monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

This paper develops a Pontryagin Differentiable Programming (PDP)methodology, which establishes a unified framework to solve a broad class oflearning and control tasks. The PDP distinguishes from existing methods by twonovel techniques: first, we differentiate through Pontryagin's MaximumPrinciple, and this allows to obtain the analytical derivative of a trajectorywith respect to tunable parameters within an optimal control system, enablingend-to-end learning of dynamics, policies, or/and control objective functions;and second, we propose an auxiliary control system in the backward pass of thePDP framework, and the output of this auxiliary control system is theanalytical derivative of the original system's trajectory with respect to theparameters, which can be iteratively solved using standard control tools. Weinvestigate three learning modes of the PDP: inverse reinforcement learning,system identification, and control/planning. We demonstrate the capability ofthe PDP in each learning mode on different high-dimensional systems, includingmulti-link robot arm, 6-DoF maneuvering quadrotor, and 6-DoF rocket poweredlanding.

Pontryagin Differentiable Programming: An End-to-End Learning and Control Framework