Modeling and optimizing flow networks with several constrains using sequential dynamical systems

Abstract

"This paper introduces a novel framework for modeling and optimizing flow networks with multiple constraints using Sequential Dynamical Systems (SDS). We have extended the Sequential Flow Network (SFN) definition to encompass both Sequential Flow Networks (SFN) and Bounded Sequential Flow Networks (bSFN), which incorporate directional constraints and weighted transitions. These models can be utilized to simulate intricate real-world applications, such as educational pathways and labor market issues. In order to optimize local flow to specific nodes with minimal global impact, we propose three novel approaches: a linear programming formulation and two greedy heuristics. The evaluation metrics employed are defined as a means to balance local improvement and global disturbance. The efficacy of these methods is evaluated through experimentation on both artificial and real-world-inspired random networks. Notwithstanding the encouraging results and observations yielded by the experimental analysis of random graphs, suggestions for further research will be made in order to overcome the limitations of the present study." (Authors' abstract, BIBB-Doku)