Instance: agricola

1. Contents:
2. Information
3. File downloads
4. Solution files
5. References

Information:

Description:

Strategic / Tactical model of an artificial, yet realistic, underground mine planning model. This problem is a large scale-RCPSP that seeks to maximize the NPV of a project, and where activities are optional. See references for model formulation and details.

Activity model:

Number of activities: 14,160

Number of arcs: 70,800

Number of time periods: 14,400 days (approx. 40 years)

Blockfile columns detail:

<activityid> <objfunction> <duration> <coeff1> <coeff2> < coeff3> <coeff4>

Objective function computation

Column objfunction gives undiscounted profit/cost associated to starting each activity

Discount rate = 0.0002 (daily)

Constraints:

Renewable knapsack constraints limit the resources that can be consumed each period. The coefficient indicates the amount of resource consumed by each activity each time period. Resources are consumed by each activity for each period in which they are being executed (as given in column duration). It is assumed that the right-hand side (resource availability) is the same for all time periods.

C0 Capacity constraint : Coefficients obtained from column coeff1. Right hand side : 6,000.0

C1 Capacity constraint : Coefficients obtained from column coeff2. Right hand side : 3,200.0

C2 Capacity constraint : Coefficients obtained from column coeff3. Right hand side : 5.0

C3 Capacity constraint : Coefficients obtained from column coeff4. Right hand side : 70.0

File downloads:

• .blocks file: AgricolaTRL.blocks.zip
  
• .prec file: AgricolaTRL.prec.zip

Best known solution files:

Feasible solution for 7200 time periods

Download: agricola7200.sol.zip

Objective value: 2,433,894,954.71

LP relaxation solution for 7200 time periods (safe version -- see IJOC paper)

Download: agricola7200.safe.lp.sol.zip

Objective value: 2,584,653,671.54 (valid upper bound)

Feasible solution for 14400 time periods

Download: agricola14400.sol.zip

Objective value: 2,482,795,122.03

References:

• Alessandro Hill, Andrea Brickey, Italo Cipriano, Marcos Goycoolea and Alexandra Newman (2022), Optimization Strategies for Resouce-Constrained Project Scheduling Problems in Underground Mining. INFORMS Journal on Computing, Accepted.
  
• Gonzalo Muñoz, Daniel Espinoza, Marcos Goycoolea, Eduardo Moreno, Maurice Queyranne, Orlando Rivera (2018), A study of the Bienstock-Zuckerberg Algorithm, Applications in Mining and Resource Constrained Project Scheduling. Computational Optimization and Applications 69(2): 501-534