Ex_10.pdf

(196 KB) Pobierz
EXERCISE 10 – SimRunner
Up to now we in-cooperated a try and error method to determine the optimum frequency. This
method is suitable for simple systems, but becomes infinite more complex once a multitude of
parameters needs to be optimized. For more advanced optimisation ProModel includes the
optimization tool: SimRunner.
SimRunner is an integrated optimisation and analysis tool for models created within the
ProModel simulation software. It allows multi-criteria optimisation for finding the best solution
for an objective function by testing various combinations of input factors (which were defined
by the model designer). Its included analytical functions can determine the number of system
replications, such that a stable behaviour of the objective function can be achieved (taking into
account the margin of error, the confidence level and the total simulation time). The SimRunner
results can be presented in form of Data, Analysis and Graphical reports to be used as a decision
support.
In order to use SimRunner the to-be-analysed model has to be running; complete; valid and
contain macros (input factors), which are the base for what to optimize. Note, that not all factors
might have an influence on your system. Therefore you should take the time to establish which
factors affect the simulation results. Please refer to the SimRunner User Guide for additional
information.
Modelling Instruction:
(General Steps are explained in the SimRunner User Guide from page 32 – 39)
1. Macro – Add Frequency_Shaft and Frequency_Bearing with a default value of 10 and
defined Options for RTI in the Range from 5 to 15.
2. Arrivals – Replace the Frequency values for the Shaft and Bearing with their respective
Macros.
3. Validate that your model is complete and running, then open the SimRunner Tool
(Simulation/SimRunner)
4. Setup Project:
a. Select model/project – Create a new project based on the previous Exercise
b. Define objectives – Select the response statistics such that:
i. The utilization of Worker resource is maximum (Resurce\Worker -
%Utilization)
ii. The average time of the Shaft Entity in the system is minimum
(Variable\Average_Cycle_time - Average_value)
Be carefull and do not choose the date from Entity group (i.e. Entity\Shaft-!
The objective function is a summation of weighted minimums or maximums times
the corresponding chosen factors. Set the weight for each factor to one.
c. Define inputs – Add Frequency_Shaft and Frequency_Bearing
Analyze Model (determines the number of replications)
a. Define experiment
i. Simulation run length: 3 hr
ii. Output recording time interval (Period): 3 min
iii. Number of test replications: 5
iv. Percent error in objective function estimate: 1
5.
6.
7.
v. Confidence Level: 95
b. Conduct analysis – Run the analysis and note the No. of replications
Optimize Model
a. Set Options –
i. Perform a moderate optimization profile with a convergence of 0.01
ii. Set the Number of Replications per experiment based on the previous result
b. Seek optimum – Run the optimization
c. Response plot – View influence of values on the objective function,
Run the optimization (a couple of times) with different parameters to see their influence
e.g.:
a. Using different optimization profiles
b. Changing weights in the Objective Function
c. Changing the Objective Function eg. Minimum Total Cost, Max Utilization of
certain workstation and/or worker, Minimum Work in Progress…
d. Using different variation of Macro for the Number of Workers and/or Deburring
Capacity…

Plik z chomika:

Inne pliki z tego folderu:

Inne foldery tego chomika:

Zgłoś jeśli naruszono regulamin