Underground Scheduling

download Underground Scheduling

of 18

Transcript of Underground Scheduling

  • 8/13/2019 Underground Scheduling

    1/18

    A MIP model for underground mine planning

    Frederic Meunier

    Work in progress with Nelson Morales

    http://find/
  • 8/13/2019 Underground Scheduling

    2/18

    Mine planning is a core discipline in mining. It starts with theresources and generates a production plan (a bankabledocument).Mine planning is hard, therefore it has been split into several

    steps.Dening what to mine (distinguishing reserves from resources)is decoupled from determining how to do it (not onlyconceptually, but often done by different teams/consultants).

    http://find/http://goback/
  • 8/13/2019 Underground Scheduling

    3/18

  • 8/13/2019 Underground Scheduling

    4/18

    Scheduling the tasks: a matter of prot

    There is a need for a tool providing a schedule of the tasks thatmaximizes the prot.

    This schedule has to be at a strategic level : time is discretized inperiods t = 1 , . . . , T .

    A period (typically: about 30 days). (Typical number of periods:55).

    For each period t:

    the prot of performing a unit of task i: vti .the availability of resource r: Ktr .

    http://find/http://goback/
  • 8/13/2019 Underground Scheduling

    5/18

    Notations

    V: set of tasks

    A: precedence constraints between tasks

    T periods (t = 1 , . . . , T ) (period t takes Dt days)

    R: set of resource types (for instance: tunnel boring machines,crushers, water pumps, workers...)

    Ktr: set of resources of type r available during period t

    1% of task i needs cri days of use of a type r resource(parallelization is allowed).

    http://find/
  • 8/13/2019 Underground Scheduling

    6/18

    A rst MIP formulation

    variables sti: equal to 1 if task i has started by period t and 0 if not

    variables eti : equal to 1 if task i has not yet ended by period t and0 if notvariables pti : percentage of task i developed during the period t

    http://find/
  • 8/13/2019 Underground Scheduling

    7/18

    A rst MIP formulation

    max i, t vti pti

    s.t. sti st+1i for all i V, t

    eti et 1i for all i V, t

    pti s

    ti

    for all

    i V, tpti eti for all i V, t

    1 eti s t psi for all i V, tsti , e

    ti {0, 1} for all i V, t

    pti 0 for all i V, t

    st j 1 et 1i for all (i, j) A, t

    i cripti | K

    tr |D

    tfor all r R, t

    http://find/
  • 8/13/2019 Underground Scheduling

    8/18

    This model has a drawback

    The constraint

    st j 1 et 1i for all (i, j) A, t = 2 , . . . , T

    prohibits feasible schedule: if task i ends at the beginning of periodt , the remaining time could be used to start task j.

    http://find/http://goback/
  • 8/13/2019 Underground Scheduling

    9/18

    A less strict constraint

    st j 1 eti for all (i, j) A, t = 1 , . . . , T

    may allow unfeasible schedule: task i may nish at the end of period t and task j at its beginning.

    http://find/http://goback/
  • 8/13/2019 Underground Scheduling

    10/18

    Morales, Rocher and Rubio (2011) proposed a way to partiallysettle the problem and succeeded in allowing more feasiblesolutions. However, their model needs an exponential number of constraints and some feasible schedules are still not covered.

    http://find/
  • 8/13/2019 Underground Scheduling

    11/18

    The one period problem

    V: set of tasks

    A: precedence constraints between tasks

    D days available

    R: set of ressource types (for instance: tunnel boring machines,crushers, water pumps, workers...)

    Kr : set of resources of type r

    task i needs cri days of use of a type r ressource (parallelization isallowed).

    Is it possible to achieve all tasks in D days ?

    http://find/
  • 8/13/2019 Underground Scheduling

    12/18

    The one period problem: necessary condition

    Assume that we have a feasible schedule and denote: i the number of days devoted to task i dri the cumulated number of days a resource of type r is

    assigned to task i i the number of days the resource is assigned to task i

    Then we have

    iV i D for all resources

    dri = Kr

    i for all tasks i and resource types r

    i maxrR, Kr i for all tasks i

    iP

    i D for any path P in the digraph (V, A)

    dri = c ri for all tasks i and resource types r

    http://find/http://goback/
  • 8/13/2019 Underground Scheduling

    13/18

    The one period problem: sufficient condition

    However, it is not a sufficient condition (counter-examples exist).Nevertheless if

    iV

    i D for all resources

    dri = Kr

    i for all tasks i and resource types r

    i = maxrR

    dri for all tasks i

    iP i D for any path P in the digraph (V, A)

    dri = c ri for all tasks i and resource types r

    then there exists a feasible schedule.

    http://find/http://goback/
  • 8/13/2019 Underground Scheduling

    14/18

    iP i D for any path P in the digraph (V, A) requires anexponential description.

    Introducing variables yi for i V reduces the number of inequalities:

    yi = 0 if i has no predecessoryi D if i has no successory j yi + i for all (i, j) A

    http://find/
  • 8/13/2019 Underground Scheduling

    15/18

    The one period problem: sufficient condition compactversion

    PropositionIf there exist yi for all i V such that

    iV cri | Kr |D for each resource type ryi = 0 if i has no predecessor

    yi D if

    i has no successor

    y j yi + max rR cri for all (i, j) A

    then there is a feasible scheduling.

    This condition can be tested through linear programming.More efficiently, through dynamic programming: once

    iV cri | Kr |D for each resource type r we can use a PERTapproach:put on each arc (i, j) a weight = max rR cri;

    feasible schedule: iff the longest path in the graph has length D.

    http://find/
  • 8/13/2019 Underground Scheduling

    16/18

    Back to the multi-period problem

    The constraints are added:iV ctri | Kr |Dt for each resource type r.

    yi = 0 if i has no predecessoryi D if i has no successory j yi + max rR ctri for all (i, j) A

    http://goforward/http://find/http://goback/
  • 8/13/2019 Underground Scheduling

    17/18

    Back to the multi-period problem: an advanced MIPformulation

    We get a new model (but still with feasible solutions not covered)

    max i, t vti pti

    s.t. usual constraintsdening

    sti , eti and sti

    st j 1 eti for all (i, j) A, t

    yi = 0 if i has no predecessoryi Dt if i has no successory j yi + max rR ctri for all (i, j) A

    i cripti | K

    tr |D

    tfor all r R, t

    http://find/
  • 8/13/2019 Underground Scheduling

    18/18

    Open questions

    For the one-period problem, is it possible to characterize feasible

    instances ?Is there even a polynomial algorithm ?

    A positive answer would lead to a compact MIP formulation of themulti-period problem.

    http://find/