Modeling and optimization of manufacturing process performance using Modelica graphical representation and process analytics formalism

Abstract

This paper concerns the development of a design methodology and its demonstration through a prototype system for performance modeling and optimization of manufacturing processes. The design methodology uses a Modelica simulation tool serving as the graphical user interface for manufacturing domain users such as process engineers to formulate their problems. The Process Analytics Formalism, developed at the National Institute of Standards and Technology, serves as a bridge between the Modelica classes and a commercial optimization solver. The prototype system includes (1) manufacturing model components’ libraries created by using Modelica and the Process Analytics Formalism, and (2) a translator of the Modelica classes to Process Analytics Formalism, which are then compiled to mathematical programming models and solved using an optimization solver. This paper provides an experiment toward the goal of enabling manufacturing users to intuitively formulate process performance models, solve problems using optimization-based methods, and automatically get actionable recommendations.

Implementing Digital Twins of Smart Factories with Interval Algebra

“…In contrast, an advantage of analytical techniques is the ease of computing using explicit mathematical formulas and numerical computation methods [19]. As observed by Shao et al [20], simulation-based evaluations are customised for special purposes and are difficult to apply them to other scenarios. Although all these drawbacks are significant, applying analytical techniques as alternative way of performance evaluation seems to be far less popular in manufacturing optimisation.…”