Multilevel ecosystems for adaptive agents: from modelling to practical applications
DOI:
https://doi.org/10.26906/SUNZ.2025.1.87-90Keywords:
adaptive agent, multi-level ecosystem, reinforcement learningAbstract
The article considers the development of multilevel ecosystems that provide training for autonomous systems
in complex and changing conditions. The main goal of the research is to create environments with different levels of complexity,
which allows autonomous agents to adapt to dynamic scenarios, overcome obstacles and effectively use resources. This approach
is aimed at improving training methods that can be used to prepare artificial intelligence systems for work in real conditions. The
work focuses on modelling multilevel ecosystems that include static and dynamic obstacles, competitive interactions between
agents and limited resources. For this purpose, the use of reinforcement learning algorithms is proposed, which allow agents to
optimize their behavioural strategies in a constantly changing environment. The developed models contribute to a better understanding of how agents can adapt to complex conditions and what factors affect the effectiveness of their behaviour. Special attention is paid to the analysis of the possibilities of practical application of such approaches in robotics. In particular, multi-level
ecosystem models can be used to train autonomous robots operating in complex environments, for example, during rescue operations, exploration of unexplored territories or performing tasks in urban conditions. These models allow creating more adaptive
and reliable autonomous systems that are able to effectively respond to changing environmental factors. The results of the study
show that multi-level ecosystems are an effective tool for preparing autonomous systems for operation in real conditions. The
proposed approaches not only contribute to increasing the adaptability and efficiency of agents, but also open up new opportunities
for their application in various industries, including industry, logistics and scientific research. The use of multi-level environments
provides autonomous systems with an advantage in complex and unpredictable conditions, which is an urgent task of modern
science and technology.
Downloads
References
1. Lawrence Perko. Differential Equations and Dynamical Systems: Springer, 2021. – 553 p.
2. David Betounes. Differential Equations: Theory and Applications: Springer, 2020. – 634 p.
3. Paul A. Gagniuc. Markov Chains: From Theory to Implementation and Experimentation: Wiley, 2021. – 256 p.
4. Pierre Brémaud. Discrete Probability Models and Methods: Springer, 2020. – 556 p.
5. Sebastian Risi, Julian Togelius. Neuro evolution: From Algorithms to Applications: Springer, 2020. – 150 p.
6. Kenneth O. Stanley, Joel Lehman. Why Greatness Cannot Be Planned: The Myth of the Objective: Springer, 2020. – 141 p.
7. Kevin Leyton-Brown, Yoav Shoham. Essentials of Game Theory: A Concise, Multidisciplinary Introduction: Morgan & Claypool Publishers, 2020. – 88 p.
8. Martin J. Osborne. An Introduction to Game Theory: Oxford University Press, 2021. – 533 p.
9. Michael Frame. Fractal Worlds: Grown, Built, and Imagined: Yale University Press, 2021. – 312 p.
10. Christoph Bandt, Kenneth Falconer, Martina Zähle. Fractal Geometry and Stochastics VI: Birkhäuser, 2021. – 340 p.
11. Maria Gini, Toru Ishida, Cristiano Castelfranchi, W. Lewis Johnson. Massively Multi-Agent Systems I: Springer, 2020. – 300 p.
12. Adelinde M. Uhrmacher, Danny Weyns. Multi-Agent Systems: Simulation and Applications: CRC Press, 2020. – 566 p.
13. Myron Tribus. Thermostatics and Thermodynamics: Springer, 2021. – 698 p.
14. Arieh Ben-Naim. Entropy and the Second Law: Interpretation and Misinterpretations: World Scientific, 2020. – 250 p.
15. Maurice Clerc. Particle Swarm Optimization: Wiley-ISTE, 2020. – 243 p.
16. Daniel Bratton, James Kennedy. Computational Intelligence: The Honey Bee Swarm Approach: Springer, 2021. – 150 p.
17. David G. Luenberger, Yinyu Ye. Linear and Nonlinear Programming: Springer, 2021. – 546 p. DOI: https://doi.org/10.1007/978-3-030-85450-8
18. Stephen Boyd, Lieven Vandenberghe. Introduction to Applied Linear Algebra: Vectors, Matrices, and Least Squares: Cambridge University Press, 2020. – 477 p.
19. Richard S. Sutton, Andrew G. Barto. Reinforcement Learning: An Introduction: MIT Press, 2020. – 552 p.
20. Marco Wiering, Martijn van Otterlo. Reinforcement Learning: State-of-the-Art: Springer, 2020. – 638 p
Downloads
Published
Issue
Section
License
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
