Application of machine learning in the process of commander decision support in the military fuel distribution system
DOI:
https://doi.org/10.24425/ather.2025.156850Abstract
Providing energy to troops requires maintaining optimal fuel levels across all management stages, especially within Garrison Support Units and Regional Logistic Bases. The article examines the fuel distribution system supported by a program that predicts commanders’ actions using input data from subordinate units. To aid decision-making, Garrison Support Units implemented neural network variants to model logistical activities, training, peacetime operations, or combat, and segment fuel supply accordingly. The Neural Network Toolbox from MATLAB (MathWorks) was used for computations. The study presents the Garrison Support Units operational assumptions, the role of commanders as agents, and factors affecting fuel distribution. It also outlines the development of the Logistic Decision Support System dashboard, which enables entering decision variables, neural network coefficients, and weights to forecast fuel consumption and plan future operations based on environmental and operational data. The article includes MATLAB simulation results, analysing neural network algorithms and neuron counts per layer to determine the most effective configuration for decision-making optimisation. Results show that the Bayesian regularisation algorithm achieved the lowest mean square error across all data sets and the highest prediction accuracy measured by the root mean squared error. The regression coefficient confirmed a strong correlation between predicted and actual outcomes, demonstrating the Bayesian regularisation algorithm’s effectiveness in supporting logistical fuel management decisions.References
[1] Kępczyński, A. (2024). A model of supporting the decision-making process for the needs of providing energy to the Armed Forces. Perspectives of Military and Civilian Logistics in the Context of Ensuring Resilience to the Armed Force, 12–14 June, Bydgoszcz, Poland (in Polish).
[2] Prokopowicz, W., & Kępczyński, A. (2024). Application of neural networks to model decision-making processes in a military fuel distribution system. Perspectives of Military and Civilian Logistics in the Context of Ensuring Resilience to the Armed Force, 12–14 June, Bydgoszcz, Poland (in Polish).
[3] Kępczyński, A., & Stępień, S. (2024). Optimization and AI support in military logistics. Perspectives of Military and Civilian Logistics in the Context of Ensuring Resilience to the Armed Force, 12–14 June, Bydgoszcz, Poland (in Polish).
[4] Kępczyński, A., Lisowski, G., Prokopowicz, W., & Stępień, S. (2023). Adaptive methods in the logistics of military energy. In: Science for Defense and the Environment (pp. 20–28). ITWL Publishing House, Warsaw (in Polish).
[5] Alasmari, T., & Alzahrani, A. (2025). Saudi Arabia’s shifts towards green and sustainable logistics: Bibliometric and machine learning-based insights and forecasts. Journal of Cleaner Production, 509, 145577. doi: 10.1016/j.jclepro.2025.145577
[6] Eichenseer, P., Hans, L., & Winkler, H. (2024). A data-driven machine learning model for forecasting delivery positions in logistics for workforce planning. Smart Computing and Applications, 9, 100099. doi: 10.1016/j.sca.2024.100099
[7] Nguyen, S., Gadel, M., Wang, K., Li, J., Zhang, X., Kong, S.-C., Fu, X., & Qin, Z. (2025). Maritime decarbonization through machine learning: A critical systematic review of fuel and power prediction models. Cleaner Logistics and Supply Chain, 14, 100210. doi: 10.1016/j.clscn.2025.100210
[8] Yaiprasert, C., & Hidayanto, A.N. (2023). AI-powered ensemble machine learning to optimize cost strategies in logistics business. International Journal of Industrial and Manufacturing Engineering, 4(1), 100209. doi: 10.1016/j.jjimei.2023.100209
[9] Zhou, J., Yang, C., Liu, D., Wang, Y., Zhong, Z., & Wu, Y. (2025). A three-stage geospatial network optimal location decision model for urban green logistics centers from a sustainable perspective. Sustainable Cities and Society, 128, 106481. doi: 10.1016/j.scs.2025.106481
[10] Liang, Y.Y., Shahabuddin, M., Ahmed, S.F., Tan, J.X., & Ali, S.M. (2025). Optimizing sustainable aviation fuel supply chains: Challenges, mitigation strategies and modeling advances. Fuel, 402, 135972. doi: 10.1016/j.fuel.2025.135972
[11] Burden, F.R., & Winkler, D.A. (2008). Bayesian regularization of neural networks. Methods in Molecular Biology, 458, 25–44. doi: 10.1007/978-1-60327-101-1_3
[12] Hagan, M.T., & Menhaj, M.B. (1994). Training feedforward networks with the Marquardt algorithm. IEEE Transactions on Neural Networks, 5(6), 989–993. doi: 10.1109/72.329697
[13] Fletcher, R. (1970). A new approach to variable metric algorithms. The Computer Journal, 13(3), 317–322. doi: 10.1093/comjnl/13.3.317
[14] Riedmiller, M., & Braun, H. (1993). A direct adaptive method for faster backpropagation learning: The RPROP algorithm. IEEE International Conference on Neural Networks, 586–591. 28 March–01 April, San Francisco, USA. doi: 10.1109/ICNN.1993.298623
[15] Ficoń, K. (2006). Applied Operational Research. Models and Applications. BEL STUDIO Publishing House, Warsaw (in Polish).
[16] Saber, H.A., Al-Barwari, R.R.I., & Talabany, Z.J. (2013). Effect of ambient air temperature on specific fuel consumption of naturally aspirated diesel engine. Journal of Science and Engineering, 1(1), 1–7.
[17] Bielaczyc, P.A., Szczotka, A., & Woodburn, J. (2011). The effect of a low ambient temperature on the cold-start emissions and fuel consumption of passenger cars. Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering, 225(9), 1253–1264. doi: 10.1177/0954407011406613
[18] Andrych-Zalewska, M., Chłopek, Z., Merkisz, J., & Pielecha, J. (2023). Impact of the internal combustion engine thermal state during start-up on the exhaust emissions in the homologation test. Energies, 16(4), 1937. doi: 10.3390/en16041937
[19] Kayri, M. (2016). Predictive abilities of Bayesian regularization and Levenberg–Marquardt algorithms in artificial neural networks: A comparative empirical study on social data. Mathematical and Computational Applications, 21(2). doi: 10.3390/mca21020020
[20] Burden, F., & Winkler, D. (2008). Bayesian regularization of neural networks. In D.J. Livingstone (Ed.), Artificial Neural Networks: Methods in Molecular Biology, 458. Humana Press. doi: 10.1007/978-1-60327-101-1_3
[21] Sariev, E., & Germano, G. (2020). Bayesian regularized artificial neural networks for the estimation of the probability of default. Quantitative Finance, 20(2), 311–328. doi: 10.1080/14697688.2019.1633014
[22] Groß, A., Friedland, J., & Schwenker, F. (2008). Learning to play Tetris applying reinforcement learning methods. 16th European Symposium on Artificial Neural Networks, 23–25 April, Bruges, Belgium.
[23] Stevens, M., & Pradhan, S. (2016). Playing Tetris with deep reinforcement learning. Stanford University.
[24] Mnih, V., Kavukcuoglu, K., Silver, D., Rusu, A.A., Veness, J., Bellemare, M.G., Graves, A., Riedmiller, M., Fidjeland, A.K., Ostrovski, G., Petersen, S., Beattie, C., Sadik, A., Antonoglou, I., King, H., Kumaran, D., Wierstra, D., Legg, S., & Hassabis, D. (2015). Human-level control through deep reinforcement learning. Nature, 518(7540), 529–533. doi: 10.1038/nature14236
[25] Kaelbling, L.P., Littman, M.L., & Moore, A.W. (1996). Reinforcement learning: A survey. Journal of Artificial Intelligence Research, 4, 237–285.
