Humans' behavior is driven by needs and changing preferences, making decision-making complex. To model this for robots, we use Hull's motivation theory, simulating an agent that strives for homeostasis. By incorporating hedonic dimensions and reinforcement learning, we train robots with different energy decay rates in various environments to study how these factors affect their strategies and behavior.
Behavior learning occurs from the expected reward resulting from drive reduction.
Code: M1_mechanism.ipynb
Behavior learning occurs from the expected reward resulting from drive reduction and hedonic dimension.
Code: M2_mechanism.ipynb
All the examples in this repository are distributed under a non-commercial license. If you use this environment, you have to agree with the following items:
- To cite our associated references in any of your publications that make any use of these examples.
- To use the environment for research purposes only.
- To not provide the environment to any second parties.
- Berto, L. et al. (2024) 'A motivational-based learning model for Mobile Robots', Cognitive Systems Research, 88, p. 101278. doi: 10.1016/j.cogsys.2024.101278.