202
7 Summary and Outlook
Q2. How can an energy efficiency optimization of a production system be executed without causing any restrictions on production flexibility, without
influencing the quality or the output of the production?
Firstly, all existing production requirements need to be included in the simulation model. Relevant production data, e.g., shift and maintenance plans, setup
times, and production quantities, as well as known technical restrictions, for
example minimum retention times in a machine state, batch sizes and machine
capacities, have to be modeled. Secondly, the production flow functions as a
pace maker process. The number of produced goods, the production machine
states or state durations are no subjects to optimization. It is assumed, that
production times, quality and process setups are ideal already. Thus, the optimization experiments are only focusing on non-value-adding production times
and therefore optimization proposals to increase the energy efficiency are not
at the expense of the output quantity.
The same applies for the peak load optimization. As the peaks can only be
reduced by using machine offsets of a length that does not cause a change of
the output, delays are permitted only within a time frame of a few minutes per
production shift.
Can energy be integrated as a control parameter for production optimization?
The machines with the medium and the high complexity machine logic types
contain non-productive but energy-consuming states (idle and standby state).
The non-productive states are generally used to bridge production interruptions. Depending on the length of a production stop, it is more energy efficient to
shut the machines down and restart them later than having them remain in idle
or standby for the whole duration of the interruption. It is possible to determine
the period of time that the non-value-adding states should be used for bridging
production interruptions and as of when a shutdown of the machines is the
more energy-efficient variant. To be able to perform the calculation, optimization parameters (idle and standby optimizer) were included in the machine
logic.
The simulation-based optimization methodology contains two optimization
experiments, each comprising an objective function and different optimization
parameters. The optimal parameter configurations for the objective functions
are determined during the optimization experiment runs. The total energy consumption optimization aims at finding the parameter configuration for the
7 Summary and Outlook
Q2. How can an energy efficiency optimization of a production system be executed without causing any restrictions on production flexibility, without
influencing the quality or the output of the production?
Firstly, all existing production requirements need to be included in the simulation model. Relevant production data, e.g., shift and maintenance plans, setup
times, and production quantities, as well as known technical restrictions, for
example minimum retention times in a machine state, batch sizes and machine
capacities, have to be modeled. Secondly, the production flow functions as a
pace maker process. The number of produced goods, the production machine
states or state durations are no subjects to optimization. It is assumed, that
production times, quality and process setups are ideal already. Thus, the optimization experiments are only focusing on non-value-adding production times
and therefore optimization proposals to increase the energy efficiency are not
at the expense of the output quantity.
The same applies for the peak load optimization. As the peaks can only be
reduced by using machine offsets of a length that does not cause a change of
the output, delays are permitted only within a time frame of a few minutes per
production shift.
Can energy be integrated as a control parameter for production optimization?
The machines with the medium and the high complexity machine logic types
contain non-productive but energy-consuming states (idle and standby state).
The non-productive states are generally used to bridge production interruptions. Depending on the length of a production stop, it is more energy efficient to
shut the machines down and restart them later than having them remain in idle
or standby for the whole duration of the interruption. It is possible to determine
the period of time that the non-value-adding states should be used for bridging
production interruptions and as of when a shutdown of the machines is the
more energy-efficient variant. To be able to perform the calculation, optimization parameters (idle and standby optimizer) were included in the machine
logic.
The simulation-based optimization methodology contains two optimization
experiments, each comprising an objective function and different optimization
parameters. The optimal parameter configurations for the objective functions
are determined during the optimization experiment runs. The total energy consumption optimization aims at finding the parameter configuration for the