Sökning: onr:"swepub:oai:DiVA.org:kth-337893" >
Optimizing Crop Man...
Optimizing Crop Management with Reinforcement Learning and Imitation Learning
-
- Tao, Ran (författare)
- University of Illinois at Urbana-Champaign Champaign, USA
-
- Martin, Nicolas F. (författare)
- University of Illinois at Urbana-Champaign Champaign, USA
-
- Zhao, Pan (författare)
- University of Illinois at Urbana-Champaign Champaign, USA
-
visa fler...
-
- Harrison, Matthew T. (författare)
- University of Tasmania Hobart, Australia
-
- Wu, Jing (författare)
- University of Illinois at Urbana-Champaign Champaign, USA
-
- Ferreira, Carla (författare)
- Stockholm University Stockholm, Sweden
-
- Kalantari, Zahra, 1979- (författare)
- KTH,Vatten- och miljöteknik
-
- Hovakimyan, Naira (författare)
- University of Illinois at Urbana-Champaign Champaign, USA
-
visa färre...
-
(creator_code:org_t)
- International Foundation for Autonomous Agents and Multiagent Systems (IFAAMAS), 2023
- 2023
- Engelska.
-
Ingår i: 22nd International Conference on Autonomous Agents and Multiagent Systems, AAMAS 2023. - : International Foundation for Autonomous Agents and Multiagent Systems (IFAAMAS). ; , s. 2511-2513
- Relaterad länk:
-
https://urn.kb.se/re...
Abstract
Ämnesord
Stäng
- To increase crop yield while minimizing environmental impact, we present an intelligent crop management system that optimizes nitrogen fertilization and irrigation simultaneously via reinforcement learning (RL), imitation learning (IL), and crop simulations using DSSAT. We first use deep RL to train management policies that require a large number of state variables from the simulator as observations (denoted as full observation). We then invoke IL to train management policies that only need a limited number of variables that are measurable in the real world (denoted as partial observation) by mimicking the actions of the RL-trained policies under full observation. Simulation experiments using maize in Florida demonstrate that our trained policies under both full and partial observations achieve better outcomes than a baseline policy. Most importantly, the IL-trained management policies are directly deployable in the real world as they use readily available information.
Ämnesord
- TEKNIK OCH TEKNOLOGIER -- Samhällsbyggnadsteknik -- Transportteknik och logistik (hsv//swe)
- ENGINEERING AND TECHNOLOGY -- Civil Engineering -- Transport Systems and Logistics (hsv//eng)
Nyckelord
- Imitation Learning
- Intelligent Crop Management
- Reinforcement Learning
- Sustainable Agriculture
Publikations- och innehållstyp
- ref (ämneskategori)
- kon (ämneskategori)