R.Yu. Esikov1, V.I. Budzko2, N.A. Ryndin3

1, 2 Federal Research Center «Computer Science and Control» of the Russian Academy of Sciences (Moscow, Russia)
3 Voronezh State Technical University (Voronezh, Russia)
1 vhiteroman@gmail.com; 2 Vbudzko@frccsc.ru; 3 nikitaryndin@gmail.com

This paper presents a conceptual framework for adaptive management of agricultural production systems based on the integration of Linear Programming (LP) and Reinforcement Learning (RL) methods. The approach addresses the limitations of classical LP models, which provide static, deterministic optimization under fixed conditions, by introducing a self-learning mechanism capable of adapting to stochastic environmental variations such as yield fluctuations, resource constraints, and market uncertainty. A formal environment for decision-making – CropRotationEnv – is proposed and modeled as a Markov Decision Process (MDP), where each state represents the resource status, crop history, and soil parameters, and each action corresponds to the selection of a crop for the next production cycle. The agent is trained using the Proximal Policy Optimization (PPO) algorithm, which ensures robust and stable policy updates under stochastic perturbations. The developed architecture enables dynamic optimization of crop rotation sequences, combining the analytical rigor of LP with the adaptive capabilities of RL. Theoretical analysis demonstrates that the PPO-based control mechanism provides stochastic stability, adaptability, and generalization in decision-making. The proposed approach forms the foundation for intelligent decision-support systems and digital twins of agricultural enterprises, enhancing operational resilience and efficiency in uncertain environments.

Esikov R.Yu., Budzko V.I., Ryndin N.A. Adaptive management of agricultural systems based on reinforcement learning and linear programming integration for variable activity sequences. Highly Available Systems. 2025. V. 21. № 3. P. 79−84. DOI: https://doi.org/10.18127/ j20729472-202503-07 (in Russian)

1. Schulman J. et al. Proximal policy optimization algorithms. arXiv preprint arXiv:1707.06347. 2017.
2. Fenz S. et al. AI-and data-driven crop rotation planning. Computers and electronics in agriculture. 2023. T. 212. S. 108160.
3. Goldenits G. et al. Tabular Reinforcement learning for Robust, Explainable CropRotation Policies Matching Deep Reinforcement Lear­ningPerformance. EGU General Assembly Conference Abstracts. 2024. S. 9018.
4. Goldenits G. et al. Current applications and potential future directions of reinforcement learning-based Digital Twins in agriculture. Smart Agricultural Technology. 2024. T. 8. S. 100512.
5. Budzko V.I., Medennikov V.I. E`kosistemny`j podxod k strategicheskomu upravleniyu na primere sel`skogo xozyajstva. Biznes-informatika. 2025. T. 19. № 2. S. 89–101.
6. Esikov R.Yu., Ry`ndin N.A. Upravlenie organizacionnoj sistemoj agrarnogo predpriyatiya na osnove optimizacii var`iruemoj posledovatel`nosti deyatel`nosti. Sistemy` vy`sokoj dostupnosti. 2024. T. 20. № 4. S. 44–51. DOI 10.18127/j20729472-202404-05
7. Medennikov V.I., Rajkov A.N. Analiz opy`ta cifrovoj transformacii v mire dlya sel`skogo xozyajstva Rossii. Tendencii razvitiya Internet i cifrovoj e`konomiki. Trudy` III Vseross. c mezhdunar. uchastiem nauchno-prakt. konf. Simferopol`, 2020. S. 57–62.
8. Medennikov V.I. Sistemny`j analiz cifrovy`x e`kosistem proizvodstvenny`x otraslej na primere APK. Cifrovaya e`konomika. 2021. № 3(15). S. 34–51.
9. Niklyaev V.S., Kosinskij V.S., Tkachev V.V., Suchilina A.A. Osnovy` texnologii sel`skoxozyajstvennogo proizvodstva. Zemledelie i rastenievodstvo. M.: By`lina. 2000. 391 c.
10. Evtefeev Yu.V., Kazancev G.M. Osnovy` agronomii: uchebnik dlya vuzov. M.: E`ksmo. 2013. 368 s.
11. Gerdt V.P., Lassner W. Isomorphism verification for complex and real Life algebras by Grobner basis technique / Modern Group Analysis: Advanced Analytical and Computational Methods in Mathematical Physics: Proceedings of the International Workshop Acireale. Catania, Italy. October 27–31, 1992. Springer Netherlands. 1993. S. 245–254.
12. Santos L.M.R., Michelon P.R.H., Arenales M.N., Santos R.H.S. Crop rotation scheduling with adjacency constraint. In Annals of Operations Research. 190. 2011. C. 165–180.
13. Aliano Filho A.H., Florentino de Oliveira, Vaz Pato M. Metaheuristics for a crop rotation problem. International Journal of Metaheuristics. 2014. T. 3. № 3. S. 199–222.
14. Schoning J., Richter M.L. AI-based crop rotation for sustainable agriculture worldwide. 2021 IEEE Global Humanitarian Technology Confe–rence (GHTC). IEEE. 2021. S. 142–146.
15. Ry`ndin N.A., Skvorczov Yu.S., Tishukov B.N. Cifrovizaciya upravleniya v organizacionny`x sistemax agropromy`shlenny`x predpriyatij. Pod red. A.A. Ry`ndina. Voronezh: Izdatel`sko-poligraficheskij centr «Nauchnaya kniga». 2022. 148 s.