Publication detail

VIRTUAL TRACTOR - TOOL FOR AI TESTING

PROKOP, A. HEJČL, P. REKEM, J. ŘEHÁK, K.

English title

VIRTUAL TRACTOR - TOOL FOR AI TESTING

Type

presentation, poster

Language

en

Original abstract

New technologies for autonomous operation of agricultural machinery will increase efectivity, reduce environmental impact and improve operational safety. The development of these new control algorithms needs an accurate and flexible environment for testing purposes. The presentation discusses the methodology of building such environments that contain a validated physics-based multi-body model of the agricultural tractor. The control algorithms based on reinforcement learning strategies are also explained in detail.

English abstract

New technologies for autonomous operation of agricultural machinery will increase efectivity, reduce environmental impact and improve operational safety. The development of these new control algorithms needs an accurate and flexible environment for testing purposes. The presentation discusses the methodology of building such environments that contain a validated physics-based multi-body model of the agricultural tractor. The control algorithms based on reinforcement learning strategies are also explained in detail.

Keywords in English

agricultural tractor, reinforcement learning, multi body model, autonomous driving

Released

23.06.2023

Publisher

MSC.Software s.r.o.

Location

Brno

Book

Proceedings of SOFTWARE USER CONFERENCE 2023 CZ&SK

Pages count

15