On the Study of Fuzzy Z-Graphs with Structure, Properties, and Potential Applications

Abstract

Real world networks are frequently described by information that is not only imprecise but also of varying reliability. Classical graphs represent relations crisp, while fuzzy graphs incorporate graded memberships; however, standard fuzzy graph models typically do not distinguish between the imprecision of a statement and the credibility of that statement. Motivated by Zadeh’s Z numbers, this thesis investigates a reliability aware network model called the fuzzy Z-graph. In this framework, vertices and/or edges are labeled by Z-numbers of the form Z = (A,B), where A is a fuzzy restriction describing the primary uncertain value (e.g., strength, cost, risk, or affinity) and B ∈ [0, 1] represents the reliability (confidence) associated with that restriction. We develop a rigorous mathematical foundation for fuzzy Z-graphs by providing pre- cise definitions, consistent notation, and canonical constructions for vertex labeled, edge labeled, and mixed Z-labeling schemes. Suitable aggregation operators and orderings on Z-numbers are introduced to enable comparison and optimization in Z-labeled networks. Fundamental structural properties are then established, including closure behavior under basic graph operations, well-defined Z-degree concepts under admissible aggregation, and path-based calculus with reliability compounding. Algorithmic aspects are also addressed through Z-weight evaluation and scalarization strategies that connect fuzzy Z-graphs to classical combinatorial optimization methods. Finally, the thesis outlines potential applica- tion settings such as risk reliability modeling, decision support, and uncertain infrastructure networks demonstrating how fuzzy Z-graphs offer a principled way to encode both “what is believed” and “how reliably it is believed” within a single graph-theoretic object.