Methods for Hybrid Modeling and Simulation-Based Optimization in Energy-Aware Production Planning

FBS 37 (Series 'Fortschrittsberichte Simulation / Advances in Simulation'); ISBN ebook: 978-3-903347-37-3, DOI: 10.11128/fbs.37, ARGESIM Wien 2020; ISBN print: 978-3-903311-11-4

About this Book

This thesis presents a hybrid discrete/continuous modelling approach and simulation implementation based on the Discrete-Event System Specification (DEVS) for industrial production planning optimization. The chosen hybrid modelling concept allows to accurately capture material flows within a production system as discrete entities on one hand and continuous energy flows on the other hand, in the production as well as for other domains, such as heating or cooling, up to the thermal building envelope. This enables the user to assess the energy demand during production while also incorporating dynamic interactions between production equipment and periphery. The method and implementation show structural advantages in comparison with the alternative co-simulation approach.
The author then combines this hybrid simulation with a meta-heuristic optimization procedure for operative production planning. The multi-objective problem takes into account overall energy efficiency as well as classical production goals and also optimizes production set-up times in order to save energy.
To support efficient modelling of such hybrid simulations in practical applications, the author also designs a domain-specific model abstraction level with modules, called Cubes, which allows to describe hybrid simulation models in an intuitive and domain-specific way. The method is based on a model-driven engineering process and defines a component-based meta-model for production systems. And last but not least, the modelling and simulation concept is demonstrated on a practical application of a bakery factory.

About the Author

Bernhard Heinzl studied Technical Mathematics and Automation Engineering at the Vienna University of Technology (TU Wien). His interests led him to specializing in Modelling and Simulation with focus on industrial systems. In his Diploma Thesis "Object-Oriented Multi-Domain Modelling and Simulation of Machine Tools", he already dealt with the upcoming demand for multi-domain modelling for improving the energy efficiency of industrial processes. His second Diploma Thesis with the title "Hybrid Modeling of Production Systems: Co-Simulation and DEVS-Based Approach" carefully analyses co-simulation for combining continuous and discrete simulation models. This experience offered him a position as a research assistant at the Institute of Computer Engineering, Automation Systems Group, at TU Wien, where he oriented his research away from co-simulation, towards DEVS-based methods. Within several projects on energy efficiency in industry, he also dove into research on simulation-based meta-heuristic optimization and automating model generation through model-driven engineering techniques, finally resulting in an exciting PhD thesis on employing hybrid modelling and simulation for energy-aware production scheduling in industry. Bernhard Heinzl has also been involved in teaching Modelling and Simulation, starting as tutor at TU Wien, and now giving lectures on theory of simulation at the University of Applied Sciences in Vienna. He now continues his career and focuses towards combining simulation with Data Science.