xii
Summary
production processes. To map the work steps of the individual machines as accurate as possible, an agent-based machine logic is defined, which can represent
the processes using status diagrams. This allows a realistic modeling of machine
states, state changes and state durations of machine types of varying complexity.
The connection to energy consumption behavior is made via the energy component of the simulation module. The energy component includes the modeling of
continuous energy load profiles, which are assigned to the individual machine
states and are used in the simulation runs according to the machine switching
behavior to determine the total energy consumption of the production processes.
In order not only to model the energy aspects in production but also to use
them for optimization scenarios, lexicographically ordered objective functions are
derived, which determine ideal parameter configurations for the energy-efficient
operation of the production lines in simulation-based optimization experiments.
The focus of the optimization is on the reduction of the total energy consumption by avoiding non-value-adding machine states. In these phases, the production
lines consume energy unnecessarily, but are not actively used to create value.
The overall consumption optimization shows that companies can counteract this
waste of resources by efficiently switching the machines without having to make
large financial investments in new technologies. In addition to the optimization of
the total energy demand, the methodology includes the possibility to change the
machine starts within a defined period of time in order to reduce peak loads.
The developed methodology is first implemented and refined in a fictitious
case study before it is applied to the example of an automotive supplier in industrial practice. Within the scope of these practical tests, energy data with different
resolutions is tested. The practical application of the methodology shows that it is
possible to build a hybrid simulation model for the representation of energy consumption behavior in production on the basis of historical consumption data and,
in combination with forecast figures, to very accurately represent future energy
consumption with upcoming peak loads and non-value-adding production phases.
The conducted optimization experiments thus result in proposals for action for the
energy-efficient control of machines, which in production situations similar to the
fictitious example lead to energy consumption reductions of 10% and in strongly
interlinked production lines to savings of about 6%.
Summary
production processes. To map the work steps of the individual machines as accurate as possible, an agent-based machine logic is defined, which can represent
the processes using status diagrams. This allows a realistic modeling of machine
states, state changes and state durations of machine types of varying complexity.
The connection to energy consumption behavior is made via the energy component of the simulation module. The energy component includes the modeling of
continuous energy load profiles, which are assigned to the individual machine
states and are used in the simulation runs according to the machine switching
behavior to determine the total energy consumption of the production processes.
In order not only to model the energy aspects in production but also to use
them for optimization scenarios, lexicographically ordered objective functions are
derived, which determine ideal parameter configurations for the energy-efficient
operation of the production lines in simulation-based optimization experiments.
The focus of the optimization is on the reduction of the total energy consumption by avoiding non-value-adding machine states. In these phases, the production
lines consume energy unnecessarily, but are not actively used to create value.
The overall consumption optimization shows that companies can counteract this
waste of resources by efficiently switching the machines without having to make
large financial investments in new technologies. In addition to the optimization of
the total energy demand, the methodology includes the possibility to change the
machine starts within a defined period of time in order to reduce peak loads.
The developed methodology is first implemented and refined in a fictitious
case study before it is applied to the example of an automotive supplier in industrial practice. Within the scope of these practical tests, energy data with different
resolutions is tested. The practical application of the methodology shows that it is
possible to build a hybrid simulation model for the representation of energy consumption behavior in production on the basis of historical consumption data and,
in combination with forecast figures, to very accurately represent future energy
consumption with upcoming peak loads and non-value-adding production phases.
The conducted optimization experiments thus result in proposals for action for the
energy-efficient control of machines, which in production situations similar to the
fictitious example lead to energy consumption reductions of 10% and in strongly
interlinked production lines to savings of about 6%.