128
5 Development of a Simulation-based Methodology …
reference profiles and will therefore only be discussed in the practical use case to
provide information about the ideal data depth (section 6.1.2).
The hybrid model is built up in AnyLogic using the process modeling and
the SD library as well as the agent components including state charts. Several
variables, parameters and functions are in use to capture relevant interactions
between the different model levels.
5.6.2 Simulation Model Details
The material flow through the five machines is depicted using buffer and delay
elements from the process modeling standards (process-oriented DES). As the
hybrid simulation methodology comes with a flexible calculation of the actual
delay the machine causes depending on the machine states the machine is going
through, the standard delay object is linked to a machine logic, reflecting the
complexity of the machine.
Therefore, the delay type of the delay object itself is defined to cause a delay
until a ‘stop delay function’ is called. The delay time starts when a part enters the
machine. On part enter, a ‘process part function’ which is defined in the machine
logic is called by the delay object. At this point, there is a simulation paradigm
change in the model. While the entry of the part into the machine is modeled
using the process-oriented discrete event simulation paradigm, the simulation of
the machine behavior is agent-based. Considering required warmup durations as
well as the process duration itself, the delay sums up to the duration of single
machine states until the ‘stop delay function’ is called with expiration of the
production process duration. The machine delay ends, again there is a simulation
paradigm change back to the process-oriented DES, the processed part leaves the
machine for the next production step or the finished goods buffer, and the machine
remains in an idle state until the processing of the next part starts (Figure 5.26).
The production schedule as well as shift times and production shut downs
due to maintenance processes are defined in the source element as well as in
additional functions closing the machine activity at 10 pm and starting the morning shift again at 6 am using a time dependent event to start and to end the
machine activity. The event calls a general function in the main agent that again
triggers shift preparation processes as well as shut down functions in the single
machine logic types (Figure 5.27). It is assumed that all machines can be switched
off completely during the nights and weekends and thus do not consume energy
during production free times. The model does not contain stochastic elements, as
the fictional case does not require any. However, the simulation components are
5 Development of a Simulation-based Methodology …
reference profiles and will therefore only be discussed in the practical use case to
provide information about the ideal data depth (section 6.1.2).
The hybrid model is built up in AnyLogic using the process modeling and
the SD library as well as the agent components including state charts. Several
variables, parameters and functions are in use to capture relevant interactions
between the different model levels.
5.6.2 Simulation Model Details
The material flow through the five machines is depicted using buffer and delay
elements from the process modeling standards (process-oriented DES). As the
hybrid simulation methodology comes with a flexible calculation of the actual
delay the machine causes depending on the machine states the machine is going
through, the standard delay object is linked to a machine logic, reflecting the
complexity of the machine.
Therefore, the delay type of the delay object itself is defined to cause a delay
until a ‘stop delay function’ is called. The delay time starts when a part enters the
machine. On part enter, a ‘process part function’ which is defined in the machine
logic is called by the delay object. At this point, there is a simulation paradigm
change in the model. While the entry of the part into the machine is modeled
using the process-oriented discrete event simulation paradigm, the simulation of
the machine behavior is agent-based. Considering required warmup durations as
well as the process duration itself, the delay sums up to the duration of single
machine states until the ‘stop delay function’ is called with expiration of the
production process duration. The machine delay ends, again there is a simulation
paradigm change back to the process-oriented DES, the processed part leaves the
machine for the next production step or the finished goods buffer, and the machine
remains in an idle state until the processing of the next part starts (Figure 5.26).
The production schedule as well as shift times and production shut downs
due to maintenance processes are defined in the source element as well as in
additional functions closing the machine activity at 10 pm and starting the morning shift again at 6 am using a time dependent event to start and to end the
machine activity. The event calls a general function in the main agent that again
triggers shift preparation processes as well as shut down functions in the single
machine logic types (Figure 5.27). It is assumed that all machines can be switched
off completely during the nights and weekends and thus do not consume energy
during production free times. The model does not contain stochastic elements, as
the fictional case does not require any. However, the simulation components are