A Network-based Analysis of Technology-driven and Load-driven Constraints in Production Data
Constraints lead to statistical patterns in data. The initial step of this master thesis work is to quantify the characteristics of two hypothetical types of constraints in industrial production: technology-driven constraints and loaddriven constraints. That was achieved by analyzing the statistical properties of association networks over time in large data sets from steel manufacturing. Based on these results, an abstract theoretical framework was developed to understand better the connection between each type of constraint and its statistical patterns.