STUDY OF OPTIMIZATION PROBLEMS OF COMPLEX SYSTEMS UNDER CONDITIONS OF UNCERTAINTY
DOI:
https://doi.org/10.32782/tnv-tech.2024.6.15Keywords:
optimization, uncertainty, synaptic connections, iterative relaxation, algorithm adaptation, multi-criteria, global extremeAbstract
The paper considers the actual problem of applying some approaches to optimization under conditions of uncertainty in the context of research of complex systems with many interacting factors, dynamic variables and unpredictable external factors. The main factors affecting the formation of the mathematical model of the problem are established, in particular, the attention is focused on the initial parameters of the system, the characteristics of the connections between its components, as well as the specificity of the optimization function. During the study, the evolution of system states during optimization for different values of key parameters, such as the module of synaptic connections and external displacements, was analyzed, and an approach to reducing the probability of inefficient states was substantiated. It was found out how these parameters affect the formation of locally stable states that do not correspond to the global optimum of the problem, and their influence on the quality of decisions made in such conditions was also evaluated. An optimization algorithm has been developed that takes into account the features of complex systems and is based on the sequential formation of initial data, the multi-stage iterative process of the studied system and its transition to the optimal state. The method of optimal setting of the model parameters is proposed, which ensures the adaptability of the algorithm in conditions of uncertainty and increases the probability of reaching the global extreme of the objective function. An approach to the structuring of the uncertainty conditions of the system by ranking the relevance criteria is defined, which makes it possible to take into account the multi-criteria nature of the problem and increase the efficiency of the decision-making process. Practical recommendations have been developed for choosing the optimal values of the weighting factors, setting the parameters of the connection module, and adapting the algorithm depending on the complexity of the task and the level of uncertainty.
References
Bertsimas D., & Margaritis G. (2023). Global Optimization: A Machine Learning Approach. arXiv preprint arXiv:2311.01742. https://doi.org/10.1007/s10898-024-01434-9
Rakshit P., Konar A., Bhowmik P., Goswami I., Das S., Nagar A. K., Janarthanan R. Realization of an Adaptive Memetic Algorithm Using Differential Evolution and Q-Learning: A Case Study in Multirobot Path Planning. IEEE Transactions on Systems, Man, and Cybernetics: Systems. 2013. № 4 3(4). pp. 814–831. https://doi.org/10.1109/TSMC.2012.2222371
Archetti C., Feillet D., Mor A., & Speranza M. G. (2020). Dynamic Traveling Salesman Problem with Stochastic Release Dates. European Journal of Operational Research, 280(3), 832–844. https://doi.org/10.1016/j.ejor.2019.08.011
Metelenko N.G., Kovalenko O.V., Makedon V., Merzhynskyi Y.K., Rudych A.I. (2019). Infrastructure security of formation and development of sectoral corporate clusters, Journal of Security and Sustainability Issues 9(1): 77–89. http://doi.org/10.9770/jssi.2019.9.1
Dhiman, G., Garg, M., Nagar, A., Kumar, V., & Dehghani, M. (2021). A Novel Algorithm for Global Optimization: Rat Swarm Optimizer. Journal of Ambient Intelligence and Humanized Computing, 12(8), 8457–8474. https://doi.org/10.1007/s12652-020-02580-0
Dhiman G., Singh K.K., Soni M., Nagar A., Dehghani M. MOSOA: A New Multi-Objective Seagull Optimization Algorithm. Expert Systems with Applications. 2021. №168. pp. 114150. https://doi.org/10.1016/j.eswa.2020.114150
Makedon V., Kostyshyna T., Tuzhylkina O., Stepanova L., Filippov V. Ensuring the Efficiency of Integration Processes in the International Corporate Sector on the Basis of Strategic Management. Academy of Strategic Management Journal. 2019. Volume 18(SI1). URL: https://www.abacademies.org/articles/Ensuring-the-efficiency-ofintegration-processes-in-the-international-corporate-sector-on-the-basis-of-strategicmanagement-1939-6104-18-SI-1-452.pdf
Bera S., Nagar A.K., Sriram S., Subramanian K.G. An Array P System Based on a New Variant of Pure 2D Context-Free Grammars. Theoretical Computer Science. 2023. №948. pp. 114027. https://doi.org/10.1016/j.tcs.2023.114027
Archetti, C., Cattaruzza, D., Gu, W., Ogier, M., &Semet, F. (2022). A Sequential Approach for a Multi-Commodity Two-Echelon Distribution Problem. Computers & Industrial Engineering, 163, 107797. https://doi.org/10.1016/j.cie.2021.107797
Nwankwor E., Nagar A.K., Reid D.C. Hybrid Differential Evolution and Particle Swarm Optimization for Optimal Well Placement. Computational Geosciences. 2013. №17(2). pp. 249–268. https://doi.org/10.1007/s10596-012-9328-9
