7.2 Critical Appraisal of the Methodology
201
The combination of different simulation paradigms allows the exact depiction
of different production aspects to create a holistic simulation model. Following the current state of the art for production simulation, all relevant material
flow details are modeled using the process-oriented discrete event simulation.
To be able to model the energy consumption of machines, it is required to
model the different possible machine states as well as state transitions and
allowed state changes accordingly, as the energy requirements of the individual operating states can vary greatly. The agent-based simulation provides
state graphs, which are ideal to model the operational inside of machines.
While the agent-based simulation is also following a discrete time progress,
the energy consumption behavior of machines is modeled continuously using
stock and flow elements of the system dynamics library. The individual energy
profile sections of the corresponding machine states are extracted from the
measured total energy consumption profiles of the production machines and
can be assigned to the machine states. Continuous variables for the energy
state are created by considering the ratio of the passed time in a state and the
remaining time. Thus, it is possible, to depict the energy consumption at any
point in time, even in between events of the discretely modeled material flow.
The single machine and energy states of the machines can be differentiated
according to time and optimization aspects and can thus be split up into three
groups, technically required operations, value-adding, and non-value-adding
operations. While the technically required and the value-adding operations
do not provide any optimization potential, the non-value adding but energy
consuming machine times are in focus of energy efficiency optimizations. In
preparation for the use of the simulation model for those optimizations, the
simulation module needs to be parameterized as far as possible. Applicable
restrictions imposed by the production scenario as well as required parameters, allowed deviations from a given schedule, and mandatory retention times
for single machine states need to be introduced. To have machines in a production ready state just-in-time, an automated calculation to determine the
remaining time until a machine has to be ready for production can be added
to the model in preparation to define the most efficient production line setup
in the optimization scenarios.
Building the hybrid simulation model using the three sub-models based on
different simulation paradigms together with the described model adaptations allows a realistic reproduction of production and energy consumption
data and thus forms the basis for energy efficiency optimizations such as the
optimization of the total energy consumption or peak-load avoidance.
201
The combination of different simulation paradigms allows the exact depiction
of different production aspects to create a holistic simulation model. Following the current state of the art for production simulation, all relevant material
flow details are modeled using the process-oriented discrete event simulation.
To be able to model the energy consumption of machines, it is required to
model the different possible machine states as well as state transitions and
allowed state changes accordingly, as the energy requirements of the individual operating states can vary greatly. The agent-based simulation provides
state graphs, which are ideal to model the operational inside of machines.
While the agent-based simulation is also following a discrete time progress,
the energy consumption behavior of machines is modeled continuously using
stock and flow elements of the system dynamics library. The individual energy
profile sections of the corresponding machine states are extracted from the
measured total energy consumption profiles of the production machines and
can be assigned to the machine states. Continuous variables for the energy
state are created by considering the ratio of the passed time in a state and the
remaining time. Thus, it is possible, to depict the energy consumption at any
point in time, even in between events of the discretely modeled material flow.
The single machine and energy states of the machines can be differentiated
according to time and optimization aspects and can thus be split up into three
groups, technically required operations, value-adding, and non-value-adding
operations. While the technically required and the value-adding operations
do not provide any optimization potential, the non-value adding but energy
consuming machine times are in focus of energy efficiency optimizations. In
preparation for the use of the simulation model for those optimizations, the
simulation module needs to be parameterized as far as possible. Applicable
restrictions imposed by the production scenario as well as required parameters, allowed deviations from a given schedule, and mandatory retention times
for single machine states need to be introduced. To have machines in a production ready state just-in-time, an automated calculation to determine the
remaining time until a machine has to be ready for production can be added
to the model in preparation to define the most efficient production line setup
in the optimization scenarios.
Building the hybrid simulation model using the three sub-models based on
different simulation paradigms together with the described model adaptations allows a realistic reproduction of production and energy consumption
data and thus forms the basis for energy efficiency optimizations such as the
optimization of the total energy consumption or peak-load avoidance.