Type of stacker/reclaimer
Number of yards available
Global coal traffic index
Power Plant flow percentage
Direct transhipment flow percentage
Increase in component operating degeneration
Interval between scheduled maintenance operations
Based on the design of experiments (DOE), the simulation campaign was structured according to the pre-defined ranges for the independent variables.
4 central composite designs (CCD)
Each of these designs consists of the following tests:
n0 Central points (6)
ns Star points (2 5 = 10)
nf Factorial points (25 = 32)
Therefore, the campaign required 4(48) runs, or about 16 hours of simulation. The study first focused on the behavior of all the detail outputs over average and extreme values (maximum and minimum) to then concentrate on aggregations with greater significance, identifying the most suitable type of sampling each time.
At this point polynomial meta-models were created and statistically validated that can reproduce the response surfaces of the system being studied. Each of the 4 CCDs uses a different meta-model since it reproduces different plant configurations.
The structure of the meta-models is described below:
One of the fundamental objectives is also that of determining the stability of the control variables in relation to the independent variables and their sensitivity to variations in input data.
