Improving the Performance of a Mixed-Integer Production Scheduling Model for LKAB's Iron Ore Mine, Kiruna, Sweden

Abstract

LKAB operates the Kiruna underground iron ore mine, which utilizes a mining method known as large-scale sublevel caving. To optimize production scheduling at Kiruna, we present a combined (short- and long-term) resolution model using mixed-integer programming. The model, which incorporates various operational requirements unique to sublevel caving, minimizes deviations from demand to produce a schedule containing monthly time periods. However, the resulting model is large and solution times for schedules of requisite length are excessive. To expedite solution time, we develop a decomposition-based heuristic consisting of two phases: (i) solving five subproblems, and (ii) solving a modified version of the original model based upon information gained from the subproblem solutions. We compare the performance of the heuristic to solving the original model directly on 15 datasets. On average, we find that our heuristic obtains better solutions faster than solving the original problem directly. We present various limitations to our approach and suggest possible extensions by modifying the heuristic when solving for schedules of greater length.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
May 01, 2006
Accession Number
ADA449045

Entities

People

  • Michael A. Martinez

Organizations

  • Colorado School of Mines

Tags

Communities of Interest

  • Energy and Power Technologies
  • Human Systems

DTIC Thesaurus Topics

  • Air Force
  • Algorithms
  • Computer Programming
  • Computers
  • Evolutionary Algorithms
  • Integer Programming
  • Linear Programming
  • Machines
  • Materials
  • Mathematical Programming
  • Metals
  • Operations Research
  • Optimization
  • Production Rate
  • Simplex Method
  • Theses
  • United States

Fields of Study

  • Computer science

Readers

  • Industrial Economics
  • Operations Research