Abstract :

Cyber-physical systems operate through a continuous interaction between physical processes, computational intelligence, communication networks and control mechanisms. Their behaviour is rarely fully deterministic: sensor noise, delayed communication, nonlinear dynamics, regime changes and early-stage failures create uncertainty that cannot be adequately represented by a single forecasting or control method. This paper proposes an original Residual-Regime Markov Forecast–Control Framework for cyber-physical systems. The framework brings together three modelling layers that play different roles. The ARIMA component handles the basic linear time‑series structure and keeps the model interpretable. On top of that, a machine‑learning layer works on the residuals to capture the nonlinear behaviour that ARIMA cannot. The final layer uses a Markov‑style state representation, turning the forecast errors, system signals and operating conditions into probabilistic regimes that describe how the system is likely to evolve. Unlike classical hybrid forecasting models that stop at prediction, the proposed approach links prediction to decision-making by using Markov transition probabilities, hidden-state belief updates and risk-aware policy selection. The main idea is that forecast errors are not treated only as modelling imperfections; instead, they are interpreted as early indicators of changing system regimes. A simulation-oriented evaluation design is presented for an industrial cyber-physical process with normal operation, peak load and degradation conditions. The proposed framework is expected to improve predictive maintenance, anomaly anticipation and control-policy selection by connecting statistical forecasting, data-driven correction and probabilistic decision logic in a single pipeline. The contribution of the paper lies in transforming hybrid forecasting into a regime-aware forecast–control architecture suitable for intelligent CPS monitoring and adaptive technical management.

---

Keywords :

anomaly detection, ARIMA, Cyber physical systems, Machine learning, Markov regimes, predictive control, residual modelling, risk-aware decision-making.

---

References :

1. Alanen, J., Linnosmaa, J., Malm, T., Papakonstantinou, N., Ahonen, T., Heikkilä, E., & Tiusanen, R. (2022). Hybrid ontology for safety, security, and dependability risk assessments and Security Threat Analysis (STA) method for industrial control systems. Reliability Engineering & System Safety, 220, Article 108270. https://doi.org/10.1016/j.ress.2021.108270
2. Bishop, C. M. (2006). Pattern recognition and machine learning. Springer. https://doi.org/10.1007/978-0-387-45528-0
3. Box, G. E. P., Jenkins, G. M., Reinsel, G. C., & Ljung, G. M. (2015). Time series analysis: Forecasting and control (5th ed.). Wiley. ISBN: 978-1-118-67502-1.
4. Breiman, L. (2001). Random forests. Machine Learning, 45, 5–32. https://doi.org/10.1023/A:1010933404324
5. Carvalho, T. P., Soares, F. A. A. M. N., Vita, R., Francisco, R. P., Basto, J. P., & Alcalá, S. G. S. (2019). A systematic literature review of machine learning methods applied to predictive maintenance. Computers & Industrial Engineering, 137, Article 106024. https://doi.org/10.1016/j.cie.2019.106024
6. Gao, M., Li, Z., Pang, T., Xu, H., & Chen, S. (2023). Event-based security control for Markov jump cyber–physical systems under denial-of-service attacks: A dual-mode switching strategy. Applied Sciences, 13(21), Article 11815. https://doi.org/10.3390/app132111815
7. Goodfellow, I., Bengio, Y., & Courville, A. (2016). Deep learning. MIT Press.
8. Hyndman, R. J., & Athanasopoulos, G. (2021). Forecasting: Principles and practice (3rd ed.). OTexts.
9. Khashei, M., & Bijari, M. (2011). A novel hybridization of artificial neural networks and ARIMA models for time series forecasting. Applied Soft Computing, 11(2), 2664–2675. https://doi.org/10.1016/j.asoc.2010.10.015
10. Kovtun, V., Izonin, I., & Gregus, M. (2022). The functional safety assessment of cyber-physical system operation process described by Markov chain. Scientific Reports, 12, Article 7089. https://doi.org/10.1038/s41598-022-11193-w
11. Lee, E. A., & Seshia, S. A. (2017). Introduction to embedded systems: A cyber-physical systems approach (2nd ed.). MIT Press.ISBN: 978-0-262-53381-2.
12. Nikolov, N. (2026). A KPI-based model for smart management of urban green infrastructure and ecosystem services. Forestry Ideas, 32(1), 209–222.
13. Nikolov, N. (2025). Intelligent Urban Systems and Industry 5.0: Creating Adaptive Ecosystems for SustainableEnergyand Resource Management. International Journal of Current Science Research and Review, 8(1), 103-115,DOI:https://doi.org/10.47191/ijcsrr/V8-i1-11
14. Nikolov, N. (2024). Development of an Intelligent System for Managing Energy Consumption in the Home. ComputerScience and Technologies (Technical University of Varna), ISSN 1312-3335, pp. 11-20.
15. Pai, P.-F., & Lin, C.-S. (2005). A hybrid ARIMA and support vector machines model in stock price forecasting. Omega, 33(6), 497–505. https://doi.org/10.1016/j.omega.2004.07.024
16. Rudin, C. (2019). Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead. Nature Machine Intelligence, 1, 206–215. https://doi.org/10.1038/s42256-019-0048-x
17. Zhang, G. P. (2003). Time series forecasting using a hybrid ARIMA and neural network model. Neurocomputing, 50, 159–175. https://doi.org/10.1016/S0925-2312(01)00702-0