Solution Comparision

The performance of the new scheduling algorithm was tested against the commercial tools NPV Scheduler and Whittle. In all test cases, identical parameters and pushbacks were used to make sure the difference in solutions is the results of the scheduling algorithms.

 

Mineteck Scheduler vs NPV Scheduler

Case one is a project derived from the tutorial of NPV Scheduler. It involves 2 processes and 2 elements.

Summary of the Case 1 (10 pushbacks):

NPV Scheduler New Algorithm
Mining Tonnage (tonnage) 1.16529e+008 1.16529e+008
Ore Tonnage (tonnage) 5.72875e+007 5.41710e+007
NPV 7.40544e+007 7.94103e+007

 case1

 

Mineteck Scheduler vs Whittle Milawa

 

Case two is Whittle’s Marvin project. It involves 1 process and 2 elements. Different final pits have been chosen for the test.

In case 2-1 , shell 11 12 14 16 20 25 31 are chosen as pushbacks, the final pit 31 is generated by the revenue factor 0.8.

Summary of the Case 2-1:

Whittle Milawa11 12 14 16 20 25 31 New Algorithm11 12 14 16 20 25 31
Mining Tonnage (tonnage) 4.50218e+008 4.50218e+008
Ore Tonnage (tonnage) 2.76407e+008 2.73814e+008
NPV 6.88053e+008 7.13178e+008

case2-1

In case 2-2, shell 42 is added to the pushbacks and this increases the number of the pushbacks to eight. The final pit 42 is generated with the revenue factor 1.0 and it is normally considered as the final pit for open design due to its revenue factor value.

Summary of the Case 2-2:

Whittle Milawa11 12 14 16 20 25 31 42 New Algorithm11 12 14 16 20 25 31 42
Mining Tonnage (tonnage) 5.74688e+008 4.50218e+008
Ore Tonnage (tonnage) 3.13788e+008 2.73814e+008
NPV 6.67327e+008 7.13178e+008

case2-2

 

In both test cases, the new algorithm generated schedules with improved NPV. Further discussion on the case study two is described in following section.