Handling of Strongly Correlated Logistics and Production Processes

Abstract

Random number generators are widely used to model stochastic processes in logistics and production systems. Creating truly independent random numbers is one important feature of these generators. However, actual data in real world systems is rarely independent. This paper discusses occurrences of dependencies in observed data by examining sample data for correlation structures. It will be demonstrated that independent random numbers created by common generators are not suitable to model processes with distinct dependencies.