Control of a robotic complex in a stochastic uncertain dynamic environment using Petri nets

F.V. Devyatkin, D.I. Arabadzhiev, M.A. Shereuzhev, A.I. Dyshekov

---

Abstract: The scientific novelty of this work lies in the development of an approach for integrating Bayesian filtering of sensor data and colored Petri nets, implemented for the first time in the form of a hierarchical software architecture, where posterior probabilities are mapped into a dynamic marking of the network that determines the resolution of transitions.
Aim. The study is the formalization, software implementation and experimental verification of a hierarchical control system for an industrial robotic complex under conditions of stochastic uncertainty of
a dynamic environment.
Research materials and methods. The control object is a robotic system comprising a six-link manipulator with a gripper and a video camera-based vision system. The system’s task is to move multi-colored objects (red, yellow, green, and blue cubes) from four initial positions to corresponding final positions according to a specified configuration. Colored Petri nets, which describe the parallelism of operations and resource constraints, are used to formalize the discrete-event control logic. Visual information processing is implemented using a recursive Bayesian filter, taking into account a 6×6 noise matrix and a measurement confirmation mechanism (k = 10 consecutive matches), ensuring robustness to stochastic disturbances. The software implementation is written in Python 3 using the OpenCV, NumPy, and SciPy libraries. The experimental verification was carried out in 500 simulations in the Gazebo environment and 30 full-scale tests with varying noise levels of σ = 0.05…0.2 with an assessment of the RMSE metrics, the probability of false alarms, and the execution time of the manipulation cycle.
Results. This article proposes a method for controlling a robotic manipulation system under conditions of stochastic uncertainty in a dynamic environment caused by sensor noise, data transmission delays, partial observability, and unpredictable changes in the position of objects. A stochastic model of the sensor system has been developed, ensuring stable object recognition in the presence of noise and dynamic disturbances. A control system architecture is proposed, including a data filtering module and a discreteevent decision-making layer. Experimental verification was conducted in a simulation and real-world environment using a six-link manipulator.
Conclusion. The obtained results showed a reduction in the probability of false positives to 0.024% and a 15% reduction in the execution time of manipulation operations compared to the basic deterministic approach.

Keywords: colored Petri net, robotic complex control, modeling, uncertainty, dynamic data filtering, adaptive control

For citation. Devyatkin F.V., Arabadzhiev D.I., Shereuzhev M.A., Dyshekov A.I. Control of a robotic complex in a stochastic uncertain dynamic environment using Petri nets. News of the Kabardino-Balkarian Scientific Center of RAS. 2026. Vol. 28. No. 1. Pp. 25–38. DOI: 10.35330/1991-6639-2026-28-1-25-38

©   Devyatkin F.V., Arabadzhiev D.I., Shereuzhev M.A., Dyshekov A.I., 2026

References

1. Thrun S., Burgard W., Fox D. Probabilistic robotics. MIT Press, 2005. 647 p. DOI: 10.1162/artl.2008.14.2.227
2. Lima P.U. A Bayesian approach to sensor fusion in autonomous sensor and robot networks. IEEE Instrumentation & Measurement Magazine. 2007. Vol. 10. No. 3. Pp. 22–27. DOI: 10.1109/MIM.2007.4284253
3. Park B.H., Kuc T.Y., Lee J.S. Adaptive learning control of uncertain robotic systems. International Journal of Control. 1996. Vol. 65. No. 5. Pp. 725–744. DOI: 10.1080/00207179608921719
4. Zhang Y.Y. et al. Adaptive control for robots to handle uncertainties, delays and state constraints. IFAC PapersOnLine. 2023. Vol. 56. No. 2. Pp. 1234–1240. DOI: 10.1109/TCYB.2015.2411285
5. Ziparo V.A. et al. Petri net plans: A framework for collaboration and coordination in multirobot systems. Autonomous Agents and Multi-Agent Systems. 2011. Vol. 23. No. 3. Pp. 344–383. DOI: 10.1007/s10458-010-9146-1
6. Westergaard M. Verifying parallel algorithms and programs using coloured petri nets. Transactions on Petri Nets and Other Models of Concurrency VI – Berlin, Heidelberg: Springer Berlin Heidelberg, 2012. Pp. 146–168. DOI: 10.1007/978-3-642-35179-2_7
7. Westergaard M., Maggi F.M. Modeling and verification of a protocol for operational support using coloured Petri nets. International Conference on Application and Theory of Petri Nets and Concurrency. Berlin, Heidelberg: Springer Berlin Heidelberg, 2011. Pp. 169–188. DOI: 10.1007/978-3-642-21834-7_10
8. Wang R., Kristensen L.M., Stolz V. MBT/CPN: a tool for model-based software testing of distributed systems protocols using coloured Petri nets. International Conference on Verification and Evaluation of Computer and Communication Systems. Cham: Springer International Publishing, 2018. Pp. 97–113. DOI: 10.1007/978-3-030-00359-3_7
9. Liu Q. et al. A Logic Petri net model for dynamic multi agent game decision making. Engineering Applications of Artificial Intelligence. 2024. Vol. 127. Pp. 107–120. DOI: 10.1016/j.dajour.2023.100320
10. Pająk M. Assessment of using bayesian nets and Petri nets in additive manufacturing technology selection. Procedia CIRP. 2021. Vol. 99. Pp. 123–130. DOI: 10.35784/acs-2021-01
11. Hu F., Wang W., Zhou J. et al. Petri nets based digital twin drives dual arm cooperative manipulation. Computers in Industry. 2023. Vol. 147. Art. 103880. DOI: 10.1016/j.compind.2023.103880
12. Castellano-Quero M., Castillo-López M., Fernández-Madrigal J.A. et al. A multidimensional Bayesian architecture for real time anomaly detection and recovery in mobile robot sensory systems. Engineering Applications of Artificial Intelligence. 2023. Vol. 125. Art. 106673. DOI: 10.1016/j.engappai.2023.106673
13. Zhong X., Zhong X., Hu H., Peng X. A nonparametric learning visual servoing framework for robot manipulator in unstructured environments. Neurocomputing. 2021. Vol. 452. Pp. 1–12. DOI: 10.1016/j.neucom.2021.01.02
14. Lv P., Luo G., Ma Z., Li S., Yin X. Optimal multi robot path planning for cyclic tasks using Petri nets. Control Engineering Practice. 2023. Vol. 138. Art. 105600. DOI: 10.1016/j.conengprac.2023.105600

Information about the authors

Fedor V. Devyatkin, Postgraduate Student ME7, Bauman Moscow State Technical University;
5, buld. 1, 2-nd Baumanskaya street, Moscow, 105005, Russia;
Engineer (NTR), Moscow State University of Technology “STANKIN”;
1, Vadkovsky lane, Moscow, 127055, Russia;
feodor-dev@ya.ru, ORCID: https://orcid.org/0009-0000-2639-9521, SPIN-code: 7738-5724
Denis I. Arabadzhiev, Postgraduate Student ME7, Bauman Moscow State Technical University;
5, buld. 1, 2-nd Baumanskaya street, Moscow, 105005, Russia;
Engineer (NTR), Moscow State University of Technology “STANKIN”;
1, Vadkovsky lane, Moscow, 127055, Russia;
denisarabadzhiev13@gmail.com, ORCID: https://orcid.org/0009-0000-5023-4073
Madin A. Shereuzhev, Candidate of Technical Sciences, Associate Professor, Head of Laboratory, Moscow State University of Technology “STANKIN”;
1, Vadkovsky lane, Moscow, 127055, Russia;
Associate Professor, Department of Robotic systems and Mechatronics, Bauman Moscow State Technical University;
5, build. 1, 2-nd Baumanskaya street, Moscow, 105005, Russia;
shereuzhev@gmail.com, ORCID: https://orcid.org/0000-0003-2352-992X, SPIN-code: 1734-9056
Artur I. Dyshekov, Candidate of Technical Sciences, Lead Engineer, Moscow State University of Technology “STANKIN”;
1, Vadkovsky lane, Moscow, 127055, Russia;
a.I.dyshekov@gmail.com, ORCID: https://orcid.org/0009-0002-4865-5041, SPIN-code: 1159-0974

Funding

This work was supported by the Ministry of Science and Higher Education of the Russian Federation (Project No. FSFS-2024-0012).