A Complex Model for Estimating the Probabilistic and Temporal Characteristics of Data Delivery and Processing Processes in Acoustic Recognition Systems

Relevance. Digitalization of industrial production and the transition to the concept of predictive maintenance pose the challenge for developers of monitoring systems to ensure high accuracy and timeliness of diagnostics of the state of autonomously operating equipment. Acoustic fault detection based on sound signal analysis is becoming an effective tool for non-invasive monitoring. However, the practical implementation of such systems in conditions of limited computing and network resources is associated with the problems of assessing their probability-time characteristics and selecting the best operating modes. The purpose is to develop a formalized model that makes it possible to estimate delays and the probability of timely delivery of acoustic data packets in order to identify anomalies in the operation of equipment under various parameters of the system architecture. 

Methods. A two-phase model of a queuing system with controlled access is proposed to describe the data transmission process. The paper uses analytical methods, including queuing theory, Laplace ‒ Stieltjes transformations, and Poisson flow models. 

Results. Expressions are obtained for calculating the average time of delivery and processing of acoustic data packets, as well as the probability of their timely delivery and analysis. The analysis of the influence of the system parameters on the probabilistic-temporal characteristics of the process is carried out. 

The novelty lies in the introduction of a parameterized traffic prioritization mechanism and complex modeling of the architecture of an acoustic data processing system under stochastic conditions. The proposed model takes into account the specifics of industrial applications and makes it possible to predict the behavior of the system. 

The theoretical significance is determined by the expansion of the mathematical apparatus for analyzing data transmission and processing processes based on queuing theory and stochastic modeling. 

The practical significance lies in the possibility of using the model to design and configure monitoring systems to meet the specific requirements of critical infrastructure facilities.

1. Verzun N.A., Kolbanev M.O., Salieva A.R. Multi-agent ensemble algorithm for acoustic recognition of malfunctions of autonomous technological equipment. Information and Control Systems. 2025;3:14‒24. (in Russ) DOI:10.31799/1684-8853-2025-3-14-24. EDN:VHOSZZ

2. Salieva A.R., Kolbanev M.O., Verzun N.A. Multi-Agent Ensemble Program for Acoustic Recognition of Malfunctions of Autonomous Technological Equipment. Patent RF, no. 2025662662, 05.22.2025. (in Russ) EDN:CQDGCH

3. Salieva A.R., Verzun N.A., Kolbanev M.O. Optimization of strategy in reinforcement learning algorithms in logistic decision-making systems. Proceedings of Saint Petersburg Electrotechnical University. 2025;17-3:65–77. (in Russ) DOI:10.32603/2071-8985-2025-18-3-65-77. EDN:BSRJMX

4. Nasir U., Banaras F. Monitoring of mechanical systems based on the Internet of Things in real time using acoustic and vibration data. 2025. DOI:10.13140/RG.2.2.26261.64487

5. Liu T., Mao Y., Dou H., Zhang W., Yang J., Wu P., et al. Emerging Wearable Acoustic Sensing Technologies. Advanced Science. 2025;12(6). DOI:10.1002/advs.202408653

6. Orel D., Varol A. Noise-Resistant Automatic Speech Recognition for Industrial and Urban Environments. Proceedings of the 49th Annual Conference of the IEEE Industrial Electronics Society, IECON, 16‒19 October 2023, Singapore, Singapore. IEEE; 2023. DOI:10.1109/IECON51785.2023.10312708

7. Bauer W., Baranovsky J. Comparison of Deep Recurrent Neural Networks and Bayesian Neural Networks for Detecting Electric Motor Damage Through Sound Signal Analysis. arXiv:2502.10224. 2025. DOI:10.48550/arXiv.2502.10224

8. Shi S., Zong Q. Benchmarking Machine Learning Methods for Distributed Acoustic Sensing. arXiv:2503.20681v1. 2025. DOI:10.48550/arXiv.2503.20681

9. GOST R 55263-2012. Mechanical Vibration. Evaluation of Machine Vibration by Measurements on Rotating Shafts. Moscow: Standartinform Publ.; 2014. (in Russ.)

10. GOST ISO 10816-3-2002. Mechanical Vibration. Evaluation of Machine Vibration by Measurements on Non-Rotating Parts. Moscow: Standartinform Publ.; 2014. (in Russ.)

11. GOST R 57154-2016. Тechnical Diagnostics. Monitoring of Mechanical and Heat NPP Equipment. Calculation Experimental Method. Moscow: Standartinform Publ.; 2016. (in Russ.)

12. GOST R ISO 13379-1-2015. Condition Monitoring and Diagnostics of Machines. Data Interpretation and Diagnostics Techniques. Moscow: Standartinform Publ.; 2019. (in Russ.)

13. GOST R 53564-2009. Condition Monitoring and Diagnostics of Machines. Hazardous Equipment Monitoring. Moscow: Standartinform Publ.; 2019. (in Russ.)

14. Vishnevsky V.M. Theoretical Foundations of Computer Network Design. Moscow: Technosphere Publ.; 2003. 512 p. (in Russ.)

15. Verzun N.A., Kolbanev M.O., Korshunov I.L., Mikadze S.Yu. Fundamentals of modeling information systems of multiple access. Saint Petersburg: Saint Petersburg State University of Economics Publ.; 2015. 138 p. (in Russ.) EDN:XWWFYN

16. Kleinrock L. Queueing Theory. Moscow: Mashinostroenie Publ.; 1979. 432 p. (in Russ.)

17. Zayats O., Korenevskaya M., Ilyashenko A., Mulyukha V. Prioritized Retrial Queueing Systems with Randomized Push-Out Mechanism. Informatics and Automation. 2024;23(2):325–351. (in Russ) DOI: 10.15622/ia.23.2.1. EDN:KXQKLM

18. Lin Yu.-B., Liu T.-H., Yu Tsai.-S., Chang F.-M. Queue waiting time monitoring scheme for repeated review M/G/1. Calculations. 2024;12(9):191. DOI:10.3390/computation12090191. EDN:TSVIHT

19. Kim K. Delay cycle analysis of finite-buffer M/G/1 queues and its application to the analysis of M/G/1 priority queues with finite and infinite buffers. Performance Evaluation. 2020;143:102133. DOI:10.1016/j.peva.2020.102133. EDN:HHSEXX

20. Zakharikova E.B. Symbolic-form models of queuing net in the form of “input – state – output”. Economics, Statistics, and Computer Science. Bulletin of the UMO. 2012;6:188–192. (in Russ) EDN:PJBLBR

21. Phung V.Q., Bogatyrev V.F., Karmanovskiy N.S., Le V.H. Evaluation of probabilistictemporal characteristics of a computer system with container virtualization. Scientific and Technical Journal of Information Technologies, Mechanics and Optics. 2024;24(2):249–255. (in Russ) DOI:10.17586/2226-1494-2024-24-2-249-255. EDN:NSGEMF

22. Petrov M.N., Poddubetsky M.A. Probabilistic and Temporal Characteristics of Single-Line Queuing Systems with Exponential Distribution of Service Time and Input Flow with Beta Distribution. Advances in current natural sciences. 2007;12(1):120–121. (in Russ) EDN:IJLYNJ

23. Verzun N.A., Kolbanev M.O., Salieva A.R., Egorov K.N. Changes in the Technological Process of Equipment Condition Monitoring with the Implementation of Acoustic Recognition System. Proceedings of the VIth International Scientific Conference on Neural Networks and Neurotechnologies, NeuroNT, 20‒20 June 2025, St. Petersburg, Russian Federation. IEEE; 2025. DOI:10.1109/NeuroNT66873.2025.11049965

24. Mikhaylichenko A.A. Analytical review of methods for assessing the quality of classification algorithms. The Bulletin of the Adygea State University, the series “Natural-Mathematical and Technical Sciences”. 2022;4(311):52‒59. (in Russ) DOI:10.53598/2410-3225-2022-4-311-52-59. EDN:FVSMBM