Scheduling and (Integer) Linear Programming

Transcription

1 Scheduling and (Integer) Linear Programming Christian Artigues LAAS - CNRS & Université de Toulouse, France artigues@laas.fr Master Class CPAIOR Nantes Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 1 / 78

2 Outline 1 Introduction 2 Polyhedral studies and cutting plane generation 3 (Mixed) integer programming for scheduling problems 4 Column generation 5 A few Computational results Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 2 / 78

3 Outline 1 Introduction 2 Polyhedral studies and cutting plane generation 3 (Mixed) integer programming for scheduling problems Time-indexed variables Total unimodularity Sequencing and natural date variables Flow (TSP) and natural date variables Positional date and assignment variables Hybrid formulations 4 Column generation 5 A few Computational results Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 3 / 78

4 A bit of history Since the beginning, linear programming has been used to solve scheduling problems. The military refer to their various plans or proposed schedules of training, logistical supply and deployment of combat units as a program. When I first analyzed the Air Force planning problem and saw that it could be formulated as a system of linear inequalities, I called my paper Programming in a Linear Structure (Georges Dantzig) This presentation is a (non-exhaustive) survey of (integer) linear programming formulations and valid inequalities for scheduling problems Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 4 / 78

5 A simple scheduling example One machine scheduling with release dates and deadlines 2 jobs J 1 and J 2 (p 1 = 3, p 2 = 2, r 1 = 0, r 2 = 1, d 1 = 9, d 2 = 7). 1 machine. Objective function f (S) = C 1 + C 2 = S 1 + S 2 + p 1 + p min S 1 + S 2 +p 1 + p 2 S 1 r 1 S 2 r 2 S 1 + p 1 d 1 S 2 + p 2 d 2 S 2 S 1 + p 1 S 1 S 2 + p 2 S 1, S 2 integer J 1 J 2 Ci = 8 (optimal solution) Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 5 / 78

6 The scheduling polyhedron Feasible set S The feasible set is S the set of points S R n that satisfy the constraints min S 1 + S 2 +p 1 + p 2 S 1 r 1 S 2 r 2 S 1 + p 1 d 1 S 2 + p 2 d 2 S 2 S S 2 S 1 + p 1 S 1 S 2 + p 2 S 1, S 2 integer S 1 J 1 J 2 Ci = Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 6 / 78

7 The scheduling polyhedron Convex hull conv(s) The convex{ hull of S, i.e. the smallest convex set containing S : conv(s) = x R n λi R n+, x = S i=1 λ is i, } S i=1 λ i = 1 min S 1 + S 2 +p 1 + p 2 S 1 r 1 S 2 r 2 S 1 + p 1 d 1 S 2 + p 2 d 2 S 2 S conv(s) S 2 S 1 + p 1 S 1 S 2 + p 2 S 1, S 2 integer S 1 J 1 J 2 Ci = Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 6 / 78

8 The scheduling polyhedron Optimization on S and conv(s) Let f : R n R a linear function, min f (S) = min f (S) S S S conv(s) min S 1 + S 2 +p 1 + p 2 S 1 r 1 S 2 r 2 S 1 + p 1 d 1 S 2 + p 2 d 2 S 2 S conv(s) S 2 S 1 + p 1 S 1 S 2 + p 2 S 1, S 2 integer S 1 J 1 J 2 Ci = Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 6 / 78

9 The scheduling polyhedron conv(s) is a polyhedron There exists A R m n and b R n with m finite such that conv(s) = {x R n Ax b}. Hence f (S) is a LP. min S conv(s) min S 1 + S 2 +p 1 + p 2 S 1 r 1 S 2 r 2 S 1 + p 1 d 1 S 2 + p 2 d 2 S 2 S conv(s) S 2 S 1 + p 1 S 1 S 2 + p 2 S 1, S 2 integer S 1 J 1 J 2 Ci = Find a complete description of conv(s) : hard in general Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 6 / 78

10 Definitions : valid inequalities, faces and facets Let P denote a polyhedron. A valid inequality αx β is such that S P, S verifies the inequality (P is included in the halfspace induced by the inequality) A face is the intersection of P with the hyperplane {x R n αx = β} A vertex is a face of dimension 0 A facet is a face of dimension dim(p) 1 S 2 S 1 3S 1 + 5S2 7 : a valid inequality Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 7 / 78

11 Definitions : valid inequalities, faces and facets Let P denote a polyhedron. A valid inequality αx β is such that S P, S verifies the inequality (P is included in the halfspace induced by the inequality) A face is the intersection of P with the hyperplane {x R n αx = β} A vertex is a face of dimension 0 A facet is a face of dimension dim(p) 1 S 2 S 1 P {S S 1 + 3S 2 = 6} : a 0-dimensional face (3, 1) Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 7 / 78

12 Definitions : valid inequalities, faces and facets Let P denote a polyhedron. A valid inequality αx β is such that S P, S verifies the inequality (P is included in the halfspace induced by the inequality) A face is the intersection of P with the hyperplane {x R n αx = β} A vertex is a face of dimension 0 A facet is a face of dimension dim(p) 1 S 2 S 1 P {S 2S 1 + 3S 2 = 9} : a facet Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 7 / 78

13 Definitions : valid inequalities, faces and facets Let P denote a polyhedron. A valid inequality αx β is such that S P, S verifies the inequality (P is included in the halfspace induced by the inequality) A face is the intersection of P with the hyperplane {x R n αx = β} A vertex is a face of dimension 0 A facet is a face of dimension dim(p) 1 S 2 P {S 2S 1 + 3S 2 = 9} : a facet S 1 A polyhedron is fully described by the set of all facet-inducing valid inequalities Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 7 / 78

14 (Integer) linear programming-based scheduling Given a scheduling problem and the logical description of S 1 Perform a polyhedral study : Find a complete description of conv(s) with a polynomial number of linear inequalities? The problem can be solved by LP Find a complete description of conv(s) and show it has a supermodular structure? The problem can be solved by a greedy algorithm Find a partial description of conv(s)? This gives useful valid inequalities 2 Design a (mixed)-integer programming formulation More polyhedral studies, solve with branch-and-cut, branch-and-price, branch-and-cut-and-price, heuristics,... (only partly addressed here) Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 8 / 78

15 (Integer) linear programming-based scheduling Given a scheduling problem and the logical description of S 1 Perform a polyhedral study : Find a complete description of conv(s) with a polynomial number of linear inequalities? The problem can be solved by LP Find a complete description of conv(s) and show it has a supermodular structure? The problem can be solved by a greedy algorithm Find a partial description of conv(s)? This gives useful valid inequalities 2 Design a (mixed)-integer programming formulation More polyhedral studies, solve with branch-and-cut, branch-and-price, branch-and-cut-and-price, heuristics,... (only partly addressed here) Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 8 / 78

16 (Integer) linear programming-based scheduling Given a scheduling problem and the logical description of S 1 Perform a polyhedral study : Find a complete description of conv(s) with a polynomial number of linear inequalities? The problem can be solved by LP Find a complete description of conv(s) and show it has a supermodular structure? The problem can be solved by a greedy algorithm Find a partial description of conv(s)? This gives useful valid inequalities 2 Design a (mixed)-integer programming formulation More polyhedral studies, solve with branch-and-cut, branch-and-price, branch-and-cut-and-price, heuristics,... (only partly addressed here) Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 8 / 78

17 (Integer) linear programming-based scheduling Given a scheduling problem and the logical description of S 1 Perform a polyhedral study : Find a complete description of conv(s) with a polynomial number of linear inequalities? The problem can be solved by LP Find a complete description of conv(s) and show it has a supermodular structure? The problem can be solved by a greedy algorithm Find a partial description of conv(s)? This gives useful valid inequalities 2 Design a (mixed)-integer programming formulation More polyhedral studies, solve with branch-and-cut, branch-and-price, branch-and-cut-and-price, heuristics,... (only partly addressed here) Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 8 / 78

18 (Integer) linear programming-based scheduling Given a scheduling problem and the logical description of S 1 Perform a polyhedral study : Find a complete description of conv(s) with a polynomial number of linear inequalities? The problem can be solved by LP Find a complete description of conv(s) and show it has a supermodular structure? The problem can be solved by a greedy algorithm Find a partial description of conv(s)? This gives useful valid inequalities 2 Design a (mixed)-integer programming formulation More polyhedral studies, solve with branch-and-cut, branch-and-price, branch-and-cut-and-price, heuristics,... (only partly addressed here) Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 8 / 78

19 (Integer) linear programming-based scheduling Given a scheduling problem and the logical description of S 1 Perform a polyhedral study : Find a complete description of conv(s) with a polynomial number of linear inequalities? The problem can be solved by LP Find a complete description of conv(s) and show it has a supermodular structure? The problem can be solved by a greedy algorithm Find a partial description of conv(s)? This gives useful valid inequalities 2 Design a (mixed)-integer programming formulation More polyhedral studies, solve with branch-and-cut, branch-and-price, branch-and-cut-and-price, heuristics,... (only partly addressed here) Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 8 / 78

20 Outline 1 Introduction 2 Polyhedral studies and cutting plane generation 3 (Mixed) integer programming for scheduling problems Time-indexed variables Total unimodularity Sequencing and natural date variables Flow (TSP) and natural date variables Positional date and assignment variables Hybrid formulations 4 Column generation 5 A few Computational results Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 9 / 78

21 A scheduling problem that can be solved by LP Project scheduling S PS Precedence constraints E. l ij : minimum distance between the start of job i (S i ) and the start of job j (S j ). S PS = {S R +n S j S i l ij, (i, j) E} find the shortest schedule mins n+1 S j S i l ij S 0 = 0 S j 0 (i, j) E j J i Γ 1 (j) i Γ(0) max (i,j) E x ij = x 0i = 1 i Γ(j) 1 2 l ij x ij x ji j J \ {0} x 0i 0 (longest path model) Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 10 / 78 5

22 Supermodular polyhedron : definitions [Sch96] Consider a set N = {1,..., n} and its power set 2 N (set of all subsets of N) and a supermodular set function f : 2 N R, i.e. a set function that verifies : { f ( ) = 0 f (A B) + f (A B) f (A) + f (B), A, B N P(f )={x R N i A x i f (A), A N} supermodular polyhedron There is an O(n log n) greedy algorithm to find x = argmin x P(f ) cx : if i N, c i < 0, the problem is unbounded. otherwise x B(f ). Solve the problem with the following greedy algorithm { x 1 = f ({1}) xj = f ({1,..., j}) f ({1,..., j 1}), j = 2,..., n. Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 11 / 78

23 Supermodular polyhedron : definitions [Sch96] Consider a set N = {1,..., n} and its power set 2 N (set of all subsets of N) and a supermodular set function f : 2 N R, i.e. a set function that verifies : { f ( ) = 0 f (A B) + f (A B) f (A) + f (B), A, B N P(f )={x R N i A x i f (A), A N} supermodular polyhedron There is an O(n log n) greedy algorithm to find x = argmin x P(f ) cx : if i N, c i < 0, the problem is unbounded. otherwise x B(f ). Solve the problem with the following greedy algorithm { x 1 = f ({1}) xj = f ({1,..., j}) f ({1,..., j 1}), j = 2,..., n. Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 11 / 78

24 A scheduling supermodular polyhedron [Que93] Single-machine scheduling S SM Set of feasible { schedules for a set of jobs J on a single machine :} S SM = C R J C i p i, j J, C i C j + p i C j C i + p j, i, j J, i j Queyranne [Que93] showed that conv(s SM ) = {C R J j A p jc j g(a), A J } where ( ( ) ) 2 g(a) = 1 2 j A p j + j A p2 j. These valid inequalities can be obtained by Smith s rule (WSPT). Optimum for i J w ic i is obtained by sorting jobs in non decreasing w i /p i. For A = {i 1,..., i s } J, set w i = p i for i A and w i = 0 for i J \ A Opt = p i1 p i1 + p i2 (p i1 + p i2 ) p is (p i1 + p i p is ) = g(a) Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 12 / 78

25 A scheduling supermodular polyhedron [Que93] Single-machine scheduling S SM Set of feasible { schedules for a set of jobs J on a single machine :} S SM = C R J C i p i, j J, C i C j + p i C j C i + p j, i, j J, i j Queyranne [Que93] showed that conv(s SM ) = {C R J j A p jc j g(a), A J } where ( ( ) ) 2 g(a) = 1 2 j A p j + j A p2 j. These valid inequalities can be obtained by Smith s rule (WSPT). Optimum for i J w ic i is obtained by sorting jobs in non decreasing w i /p i. For A = {i 1,..., i s } J, set w i = p i for i A and w i = 0 for i J \ A Opt = p i1 p i1 + p i2 (p i1 + p i2 ) p is (p i1 + p i p is ) = g(a) Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 12 / 78

26 A scheduling supermodular polyhedron [Que93] Single-machine scheduling S SM Set of feasible { schedules for a set of jobs J on a single machine :} S SM = C R J C i p i, j J, C i C j + p i C j C i + p j, i, j J, i j Queyranne [Que93] showed that conv(s SM ) = {C R J j A p jc j g(a), A J } where ( ( ) ) 2 g(a) = 1 2 j A p j + j A p2 j. These valid inequalities can be obtained by Smith s rule (WSPT). Optimum for i J w ic i is obtained by sorting jobs in non decreasing w i /p i. For A = {i 1,..., i s } J, set w i = p i for i A and w i = 0 for i J \ A Opt = p i1 p i1 + p i2 (p i1 + p i2 ) p is (p i1 + p i p is ) = g(a) Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 12 / 78

27 A scheduling supermodular polyhedron [Que93] Single-machine scheduling S SM Set of feasible { schedules for a set of jobs J on a single machine :} S SM = C R J C i p i, j J, C i C j + p i C j C i + p j, i, j J, i j Queyranne [Que93] showed that conv(s SM ) = {C R J j A p jc j g(a), A J } where ( ( ) ) 2 g(a) = 1 2 j A p j + j A p2 j. These valid inequalities can be obtained by Smith s rule (WSPT). Optimum for i J w ic i is obtained by sorting jobs in non decreasing w i /p i. For A = {i 1,..., i s } J, set w i = p i for i A and w i = 0 for i J \ A Opt = p i1 p i1 + p i2 (p i1 + p i2 ) p is (p i1 + p i p is ) = g(a) Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 12 / 78

28 A scheduling supermodular polyhedron [Que93] Single-machine scheduling S SM Set of feasible { schedules for a set of jobs J on a single machine :} S SM = C R J C i p i, j J, C i C j + p i C j C i + p j, i, j J, i j Queyranne [Que93] showed that conv(s SM ) = {C R J j A p jc j g(a), A J } where ( ( ) ) 2 g(a) = 1 2 j A p j + j A p2 j. These valid inequalities can be obtained by Smith s rule (WSPT). Optimum for i J w ic i is obtained by sorting jobs in non decreasing w i /p i. For A = {i 1,..., i s } J, set w i = p i for i A and w i = 0 for i J \ A Opt = p i1 p i1 + p i2 (p i1 + p i2 ) p is (p i1 + p i p is ) = g(a) Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 12 / 78

29 A scheduling supermodular polyhedron [Que93] Single-machine scheduling S SM Set of feasible { schedules for a set of jobs J on a single machine :} S SM = C R J C i p i, j J, C i C j + p i C j C i + p j, i, j J, i j Queyranne [Que93] showed that conv(s SM ) = {C R J j A p jc j g(a), A J } where ( ( ) ) 2 g(a) = 1 2 j A p j + j A p2 j. These valid inequalities can be obtained by Smith s rule (WSPT). Optimum for i J w ic i is obtained by sorting jobs in non decreasing w i /p i. For A = {i 1,..., i s } J, set w i = p i for i A and w i = 0 for i J \ A Opt = p i1 p i1 + p i2 (p i1 + p i2 ) p is (p i1 + p i p is ) = g(a) Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 12 / 78

30 A scheduling supermodular polyhedron [Que93] Each inequality j S p j C j g(s) defines a facet for conv(q). conv(s SM ) is a supermodular polyhedron and the greedy algorithm for minimizing i J w i C i coincides with the Smith s rule WSPT This rediscovers the WSPT rule but also provides valid inequalities for other (NP-hard) scheduling problems. An O(n log n) separation algorithm is available to find an inequality violated by a given vector C. More scheduling supermodular polyhedra : the convex hull of the feasible start time for unit jobs on parallel machines with nonstationary speeds [QS95] the convex hull of mean busy time vectors of preemptive schedules for jobs with release dates on a single machine [GQS + 02] Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 13 / 78

31 A scheduling supermodular polyhedron [Que93] Each inequality j S p j C j g(s) defines a facet for conv(q). conv(s SM ) is a supermodular polyhedron and the greedy algorithm for minimizing i J w i C i coincides with the Smith s rule WSPT This rediscovers the WSPT rule but also provides valid inequalities for other (NP-hard) scheduling problems. An O(n log n) separation algorithm is available to find an inequality violated by a given vector C. More scheduling supermodular polyhedra : the convex hull of the feasible start time for unit jobs on parallel machines with nonstationary speeds [QS95] the convex hull of mean busy time vectors of preemptive schedules for jobs with release dates on a single machine [GQS + 02] Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 13 / 78

32 A scheduling supermodular polyhedron [Que93] Each inequality j S p j C j g(s) defines a facet for conv(q). conv(s SM ) is a supermodular polyhedron and the greedy algorithm for minimizing i J w i C i coincides with the Smith s rule WSPT This rediscovers the WSPT rule but also provides valid inequalities for other (NP-hard) scheduling problems. An O(n log n) separation algorithm is available to find an inequality violated by a given vector C. More scheduling supermodular polyhedra : the convex hull of the feasible start time for unit jobs on parallel machines with nonstationary speeds [QS95] the convex hull of mean busy time vectors of preemptive schedules for jobs with release dates on a single machine [GQS + 02] Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 13 / 78

33 A scheduling supermodular polyhedron [Que93] Each inequality j S p j C j g(s) defines a facet for conv(q). conv(s SM ) is a supermodular polyhedron and the greedy algorithm for minimizing i J w i C i coincides with the Smith s rule WSPT This rediscovers the WSPT rule but also provides valid inequalities for other (NP-hard) scheduling problems. An O(n log n) separation algorithm is available to find an inequality violated by a given vector C. More scheduling supermodular polyhedra : the convex hull of the feasible start time for unit jobs on parallel machines with nonstationary speeds [QS95] the convex hull of mean busy time vectors of preemptive schedules for jobs with release dates on a single machine [GQS + 02] Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 13 / 78

34 A scheduling supermodular polyhedron [Que93] Each inequality j S p j C j g(s) defines a facet for conv(q). conv(s SM ) is a supermodular polyhedron and the greedy algorithm for minimizing i J w i C i coincides with the Smith s rule WSPT This rediscovers the WSPT rule but also provides valid inequalities for other (NP-hard) scheduling problems. An O(n log n) separation algorithm is available to find an inequality violated by a given vector C. More scheduling supermodular polyhedra : the convex hull of the feasible start time for unit jobs on parallel machines with nonstationary speeds [QS95] the convex hull of mean busy time vectors of preemptive schedules for jobs with release dates on a single machine [GQS + 02] Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 13 / 78

35 Valid inequalities for NP-hard scheduling problems Single-machine scheduling with release dates S SMR Set of feasible schedules for a set of jobs J on a single machine with release dates { : } S SMR = S R J S i r i, j J, S i S j + l ij S j S i + l ji, i, j J, i j The problem of minimizing S S w is i is NP-hard = no chance to have a complete characterization of conv(s SMR ) However, Balas [Bal85] studied P = conv(s) and derived facet-defining inequalities. He showed that a facet-defining inequality for a subset of jobs K J induces also a facet for J. i J, S i r i induces a facet of P i, j J, i j, (l ij + r i r j )S i + (l ji + r j r i )S j d ij d ji + L i d ji + L j d ij induces a facet of P if and only if d ji < r j r i < d ij Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 14 / 78

36 Valid inequalities for NP-hard scheduling problems Single-machine scheduling with release dates S SMR Set of feasible schedules for a set of jobs J on a single machine with release dates { : } S SMR = S R J S i r i, j J, S i S j + l ij S j S i + l ji, i, j J, i j The problem of minimizing S S w is i is NP-hard = no chance to have a complete characterization of conv(s SMR ) However, Balas [Bal85] studied P = conv(s) and derived facet-defining inequalities. He showed that a facet-defining inequality for a subset of jobs K J induces also a facet for J. i J, S i r i induces a facet of P i, j J, i j, (l ij + r i r j )S i + (l ji + r j r i )S j d ij d ji + L i d ji + L j d ij induces a facet of P if and only if d ji < r j r i < d ij Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 14 / 78

37 Valid inequalities for NP-hard scheduling problems Single-machine scheduling with release dates S SMR Set of feasible schedules for a set of jobs J on a single machine with release dates { : } S SMR = S R J S i r i, j J, S i S j + l ij S j S i + l ji, i, j J, i j The problem of minimizing S S w is i is NP-hard = no chance to have a complete characterization of conv(s SMR ) However, Balas [Bal85] studied P = conv(s) and derived facet-defining inequalities. He showed that a facet-defining inequality for a subset of jobs K J induces also a facet for J. i J, S i r i induces a facet of P i, j J, i j, (l ij + r i r j )S i + (l ji + r j r i )S j d ij d ji + L i d ji + L j d ij induces a facet of P if and only if d ji < r j r i < d ij Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 14 / 78

38 Valid inequalities for NP-hard scheduling problems Single-machine scheduling with release dates S SMR Set of feasible schedules for a set of jobs J on a single machine with release dates { : } S SMR = S R J S i r i, j J, S i S j + l ij S j S i + l ji, i, j J, i j The problem of minimizing S S w is i is NP-hard = no chance to have a complete characterization of conv(s SMR ) However, Balas [Bal85] studied P = conv(s) and derived facet-defining inequalities. He showed that a facet-defining inequality for a subset of jobs K J induces also a facet for J. i J, S i r i induces a facet of P i, j J, i j, (l ij + r i r j )S i + (l ji + r j r i )S j d ij d ji + L i d ji + L j d ij induces a facet of P if and only if d ji < r j r i < d ij Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 14 / 78

39 Cutting plane generation for NP-hard scheduling problems From a partial descrition of conv(s) (relaxation), iteratively solve the separation problem for families of valid inequalities. S 2 S 1 Optimum has been found after adding Balas inequality (l ij + r i r j )S i + (l ji + r j r i )S j d ij d ji + L i d ji + L j d ij Separation is NP-hard in general = the optimum cannot be found quickly by pure cutting plane generation. Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 15 / 78

40 Cutting plane generation for NP-hard scheduling problems From a partial descrition of conv(s) (relaxation), iteratively solve the separation problem for families of valid inequalities. S 2 S 1 3S 1 + 5S2 7 Optimum has been found after adding Balas inequality (l ij + r i r j )S i + (l ji + r j r i )S j d ij d ji + L i d ji + L j d ij Separation is NP-hard in general = the optimum cannot be found quickly by pure cutting plane generation. Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 15 / 78

41 Cutting plane generation for NP-hard scheduling problems From a partial descrition of conv(s) (relaxation), iteratively solve the separation problem for families of valid inequalities. S 2 S 1 S 1 + 3S 2 6 Optimum has been found after adding Balas inequality (l ij + r i r j )S i + (l ji + r j r i )S j d ij d ji + L i d ji + L j d ij Separation is NP-hard in general = the optimum cannot be found quickly by pure cutting plane generation. Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 15 / 78

42 Cutting plane generation for NP-hard scheduling problems From a partial descrition of conv(s) (relaxation), iteratively solve the separation problem for families of valid inequalities. S 2 S 1 2S 1 + 3S 2 9 Optimum has been found after adding Balas inequality (l ij + r i r j )S i + (l ji + r j r i )S j d ij d ji + L i d ji + L j d ij Separation is NP-hard in general = the optimum cannot be found quickly by pure cutting plane generation. Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 15 / 78

43 Cutting plane generation for NP-hard scheduling problems From a partial descrition of conv(s) (relaxation), iteratively solve the separation problem for families of valid inequalities. S 2 S 1 2S 1 + 3S 2 9 Optimum has been found after adding Balas inequality (l ij + r i r j )S i + (l ji + r j r i )S j d ij d ji + L i d ji + L j d ij Separation is NP-hard in general = the optimum cannot be found quickly by pure cutting plane generation. Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 15 / 78

44 Outline 1 Introduction 2 Polyhedral studies and cutting plane generation 3 (Mixed) integer programming for scheduling problems Time-indexed variables Total unimodularity Sequencing and natural date variables Flow (TSP) and natural date variables Positional date and assignment variables Hybrid formulations 4 Column generation 5 A few Computational results Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 16 / 78

45 Principle of (mixed)-integer programming Design a good MIP formulation for the scheduling problem Solve by branch-and-bound S 2 S S 1 Remark : once x is fixed, extreme points are integral = no need for integer constraints on S Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 17 / 78

46 Principle of (mixed)-integer programming Design a good MIP formulation for the scheduling problem Solve by branch-and-bound min S 1 + S 2 +p 1 + p 2 S 1 r 1 S 2 r 2 S 1 + p 1 d 1 S 2 + p 2 d 2 S 2 S 1 + Mx p 1 S 2 S S 1 S2 + M(1 x) p 2 S 1, S 2 integer x {0, 1} S 1 Remark : once x is fixed, extreme points are integral = no need for integer constraints on S Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 17 / 78

47 Principle of (mixed)-integer programming Design a good MIP formulation for the scheduling problem Solve by branch-and-bound min S 1 + S 2 +p 1 + p 2 S 1 r 1 S 2 r 2 S 1 + p 1 d 1 S 2 + p 2 d 2 S 2 S 1 + Mx p 1 S 2 S S 1 S2 + M(1 x) p 2 S 1, S 2 integer x {0, 1} S 1 Remark : once x is fixed, extreme points are integral = no need for integer constraints on S Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 17 / 78

48 Principle of (mixed)-integer programming Design a good MIP formulation for the scheduling problem Solve by branch-and-bound min S 1 + S 2 +p 1 + p 2 S 1 r 1 S 2 r 2 S 1 + p 1 d 1 S 2 + p 2 d 2 S 2 S 1 + Mx p 1 S 2 S S 1 S2 + M(1 x) p 2 S 1, S 2 integer x {0, 1} Left node x = 1 S 1 Remark : once x is fixed, extreme points are integral = no need for integer constraints on S Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 17 / 78

49 Principle of (mixed)-integer programming Design a good MIP formulation for the scheduling problem Solve by branch-and-bound min S 1 + S 2 +p 1 + p 2 S 1 r 1 S 2 r 2 S 1 + p 1 d 1 S 2 + p 2 d 2 S 2 S 1 + Mx p 1 S 2 S S 1 S2 + M(1 x) p 2 S 1, S 2 integer x {0, 1} Right node x = 0 J 1 J 2 Ci = S 1 Remark : once x is fixed, extreme points are integral = no need for integer constraints on S Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 17 / 78

50 Case study : the resource-constrained project scheduling problem Resource-constrained project scheduling with irregular starting time costs n jobs (set J ) with integer start times S i T = {0, 1,..., T }. Precedence constraints E such that (i, j) E = S j S i + l ij. m resources (set R). Constant availability B k, k R. For each job i J : duration p i and resource requirements b ik, k R. Resource constraints i J S i t S i +p i 1 b ik B k, t T, k R. Cost function w j : {1,..., T } R. Find a schedule that minimizes i J w j(s j ). Remark 1 : R = 1, B 1 = 1, and b i1 = 1, i J Remark 2 : R = 1, B 1 2 and b i1 = 1, i J = 1-machine problem. = parallel machine problem. Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 18 / 78

51 Case study : the resource-constrained project scheduling problem Resource-constrained project scheduling with irregular starting time costs n jobs (set J ) with integer start times S i T = {0, 1,..., T }. Precedence constraints E such that (i, j) E = S j S i + l ij. m resources (set R). Constant availability B k, k R. For each job i J : duration p i and resource requirements b ik, k R. Resource constraints i J S i t S i +p i 1 b ik B k, t T, k R. Cost function w j : {1,..., T } R. Find a schedule that minimizes i J w j(s j ). Remark 1 : R = 1, B 1 = 1, and b i1 = 1, i J Remark 2 : R = 1, B 1 2 and b i1 = 1, i J = 1-machine problem. = parallel machine problem. Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 18 / 78

52 Case study : the resource-constrained project scheduling problem Resource-constrained project scheduling with irregular starting time costs n jobs (set J ) with integer start times S i T = {0, 1,..., T }. Precedence constraints E such that (i, j) E = S j S i + l ij. m resources (set R). Constant availability B k, k R. For each job i J : duration p i and resource requirements b ik, k R. Resource constraints i J S i t S i +p i 1 b ik B k, t T, k R. Cost function w j : {1,..., T } R. Find a schedule that minimizes i J w j(s j ). Remark 1 : R = 1, B 1 = 1, and b i1 = 1, i J Remark 2 : R = 1, B 1 2 and b i1 = 1, i J = 1-machine problem. = parallel machine problem. Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 18 / 78

53 Case study : the resource-constrained project scheduling problem Resource-constrained project scheduling with irregular starting time costs n jobs (set J ) with integer start times S i T = {0, 1,..., T }. Precedence constraints E such that (i, j) E = S j S i + l ij. m resources (set R). Constant availability B k, k R. For each job i J : duration p i and resource requirements b ik, k R. Resource constraints i J S i t S i +p i 1 b ik B k, t T, k R. Cost function w j : {1,..., T } R. Find a schedule that minimizes i J w j(s j ). Remark 1 : R = 1, B 1 = 1, and b i1 = 1, i J Remark 2 : R = 1, B 1 2 and b i1 = 1, i J = 1-machine problem. = parallel machine problem. Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 18 / 78

54 Case study : the resource-constrained project scheduling problem Resource-constrained project scheduling with irregular starting time costs n jobs (set J ) with integer start times S i T = {0, 1,..., T }. Precedence constraints E such that (i, j) E = S j S i + l ij. m resources (set R). Constant availability B k, k R. For each job i J : duration p i and resource requirements b ik, k R. Resource constraints i J S i t S i +p i 1 b ik B k, t T, k R. Cost function w j : {1,..., T } R. Find a schedule that minimizes i J w j(s j ). Remark 1 : R = 1, B 1 = 1, and b i1 = 1, i J Remark 2 : R = 1, B 1 2 and b i1 = 1, i J = 1-machine problem. = parallel machine problem. Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 18 / 78

55 Case study : the resource-constrained project scheduling problem Resource-constrained project scheduling with irregular starting time costs n jobs (set J ) with integer start times S i T = {0, 1,..., T }. Precedence constraints E such that (i, j) E = S j S i + l ij. m resources (set R). Constant availability B k, k R. For each job i J : duration p i and resource requirements b ik, k R. Resource constraints i J S i t S i +p i 1 b ik B k, t T, k R. Cost function w j : {1,..., T } R. Find a schedule that minimizes i J w j(s j ). Remark 1 : R = 1, B 1 = 1, and b i1 = 1, i J Remark 2 : R = 1, B 1 2 and b i1 = 1, i J = 1-machine problem. = parallel machine problem. Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 18 / 78

56 Case study : the resource-constrained project scheduling problem Resource-constrained project scheduling with irregular starting time costs n jobs (set J ) with integer start times S i T = {0, 1,..., T }. Precedence constraints E such that (i, j) E = S j S i + l ij. m resources (set R). Constant availability B k, k R. For each job i J : duration p i and resource requirements b ik, k R. Resource constraints i J S i t S i +p i 1 b ik B k, t T, k R. Cost function w j : {1,..., T } R. Find a schedule that minimizes i J w j(s j ). Remark 1 : R = 1, B 1 = 1, and b i1 = 1, i J Remark 2 : R = 1, B 1 2 and b i1 = 1, i J = 1-machine problem. = parallel machine problem. Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 18 / 78

57 Case study : the resource-constrained project scheduling problem Resource-constrained project scheduling with irregular starting time costs n jobs (set J ) with integer start times S i T = {0, 1,..., T }. Precedence constraints E such that (i, j) E = S j S i + l ij. m resources (set R). Constant availability B k, k R. For each job i J : duration p i and resource requirements b ik, k R. Resource constraints i J S i t S i +p i 1 b ik B k, t T, k R. Cost function w j : {1,..., T } R. Find a schedule that minimizes i J w j(s j ). Remark 1 : R = 1, B 1 = 1, and b i1 = 1, i J Remark 2 : R = 1, B 1 2 and b i1 = 1, i J = 1-machine problem. = parallel machine problem. Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 18 / 78

58 Case study : the resource-constrained project scheduling problem Resource-constrained project scheduling with irregular starting time costs n jobs (set J ) with integer start times S i T = {0, 1,..., T }. Precedence constraints E such that (i, j) E = S j S i + l ij. m resources (set R). Constant availability B k, k R. For each job i J : duration p i and resource requirements b ik, k R. Resource constraints i J S i t S i +p i 1 b ik B k, t T, k R. Cost function w j : {1,..., T } R. Find a schedule that minimizes i J w j(s j ). Remark 1 : R = 1, B 1 = 1, and b i1 = 1, i J Remark 2 : R = 1, B 1 2 and b i1 = 1, i J = 1-machine problem. = parallel machine problem. Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 18 / 78

59 Scheduling objectives Objective min i J w j(s j ) models most standard objectives Makespan min max i J C i Let J = {1,..., n + 1} with n + 1 dummy end jobs. w i (t) = 0, i J \ {n + 1}, w n+1 (t) = t Maximum lateness min max i J C i d i Same modeling by adding arc (i, n + 1) in E with l i,n+1 = p i d i Earliness-tardiness costs min i J (α i max(0, d i C i ) + β i max(0, C i d i ) w i (t) = α i max(0, d i t p i ) + β i max(0, t + p i d i ) (weighted completion time if, in addition, d i = 0 α i = 0, i J ) Weighted number of late jobs i J w iu i { 0 if t + pi d w i (t) = i w i otherwise Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 19 / 78

60 Scheduling objectives Objective min i J w j(s j ) models most standard objectives Makespan min max i J C i Let J = {1,..., n + 1} with n + 1 dummy end jobs. w i (t) = 0, i J \ {n + 1}, w n+1 (t) = t Maximum lateness min max i J C i d i Same modeling by adding arc (i, n + 1) in E with l i,n+1 = p i d i Earliness-tardiness costs min i J (α i max(0, d i C i ) + β i max(0, C i d i ) w i (t) = α i max(0, d i t p i ) + β i max(0, t + p i d i ) (weighted completion time if, in addition, d i = 0 α i = 0, i J ) Weighted number of late jobs i J w iu i { 0 if t + pi d w i (t) = i w i otherwise Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 19 / 78

61 Scheduling objectives Objective min i J w j(s j ) models most standard objectives Makespan min max i J C i Let J = {1,..., n + 1} with n + 1 dummy end jobs. w i (t) = 0, i J \ {n + 1}, w n+1 (t) = t Maximum lateness min max i J C i d i Same modeling by adding arc (i, n + 1) in E with l i,n+1 = p i d i Earliness-tardiness costs min i J (α i max(0, d i C i ) + β i max(0, C i d i ) w i (t) = α i max(0, d i t p i ) + β i max(0, t + p i d i ) (weighted completion time if, in addition, d i = 0 α i = 0, i J ) Weighted number of late jobs i J w iu i { 0 if t + pi d w i (t) = i w i otherwise Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 19 / 78

62 Scheduling objectives Objective min i J w j(s j ) models most standard objectives Makespan min max i J C i Let J = {1,..., n + 1} with n + 1 dummy end jobs. w i (t) = 0, i J \ {n + 1}, w n+1 (t) = t Maximum lateness min max i J C i d i Same modeling by adding arc (i, n + 1) in E with l i,n+1 = p i d i Earliness-tardiness costs min i J (α i max(0, d i C i ) + β i max(0, C i d i ) w i (t) = α i max(0, d i t p i ) + β i max(0, t + p i d i ) (weighted completion time if, in addition, d i = 0 α i = 0, i J ) Weighted number of late jobs i J w iu i { 0 if t + pi d w i (t) = i w i otherwise Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 19 / 78

63 Outline 1 Introduction 2 Polyhedral studies and cutting plane generation 3 (Mixed) integer programming for scheduling problems Time-indexed variables Total unimodularity Sequencing and natural date variables Flow (TSP) and natural date variables Positional date and assignment variables Hybrid formulations 4 Column generation 5 A few Computational results Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 20 / 78

64 MILP Formulations with time-indexed variables time-indexed variables x it = 1 S i = t S i = T t=0 tx it Single machine min w j (t)x jt j J t T x jt = 1 j J t T T T tx jt tx it l ij (i, j) E t=0 t=0 t x js 1 t T j J s=t p j +1 x jt {0, 1} j J, t T Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 21 / 78

65 MILP Formulations with time-indexed variables time-indexed variables x it = 1 S i = t S i = T t=0 tx it Parallel machines min w j (t)x jt j J t T x jt = 1 j J t T T T tx jt tx it l ij (i, j) E t=0 t=0 t x js B t T j J s=t p j +1 x jt {0, 1} j J, t T Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 21 / 78

66 MILP Formulations with time-indexed variables time-indexed variables x it = 1 S i = t S i = T t=0 tx it RCPSP min w j (t)x jt j J t T x jt = 1 j J t T T T tx jt tx it l ij (i, j) E t=0 t=0 t b jk x js B k t T, k R j J s=t p j +1 x jt {0, 1} j J, t T nt variables, E precedence constraints, R T resource constraints. Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 21 / 78

67 MILP Formulations with time-indexed variables time-indexed variables x it = 1 S i = t S i = T t=0 tx it RCPSP (disaggregated precedence constraints) min w j (t)x jt j J t T x jt = 1 j J t T t+l T ij 1 x is + x js 1 (i, j) E, t T s=t s=0 t b jk x js B k t T, k R j J s=t p j +1 x jt {0, 1} j J, t T nt variables, E T precedence constraints, R T resource constraints. Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 21 / 78

68 Outline 1 Introduction 2 Polyhedral studies and cutting plane generation 3 (Mixed) integer programming for scheduling problems Time-indexed variables Total unimodularity Sequencing and natural date variables Flow (TSP) and natural date variables Positional date and assignment variables Hybrid formulations 4 Column generation 5 A few Computational results Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 22 / 78

69 Total unimodularity A matrix A is totally unimodular (TU) if and only if every square submatrix has determinant 0, 1 or -1. if A is TU, min Ax b cx has an integer optimal solution or there is no solution A is TU if (sufficient condition) rows can be partitionned into two disjoint sets B and C such that each column of A has at most two non-zero entries, each entry of A is 0, 1 or -1 if two non-zeros in a column have opposite signs, they are in the same subset of rows (both in B or both in C). if two non-zeros in a column have the same sign, there is one in B and the other one in C A is also TU if its transpose is TU. Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 23 / 78

70 Total unimodularity and scheduling Project scheduling with irregular starting time costs n jobs (set J ) with integer start times S i T = {0, 1,..., T }. Precedence constraints E such that (i, j) E = S j S i + l ij. Cost function w j : {1,..., T } R. Find a schedule that minimizes j J w j(s j ). time-indexed variable z it = 1 S i t (z it = T t=0 x it ) min (w j (t) w j (t + 1)) z jt j J t T z jt = 1 z jt z j,t+1 0 z j,t+lij z it 0 z jt {0, 1} j J j J, t T (i, j) E, t T j J, t T The matrix totally unimodular = no integer restriction needed! Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 24 / 78

71 Comparison between formulations Other IP with time-indexed variables x it = 1 S i = t (nt variables) T x is + s=t min w j (t)x jt x jt = 1 t T t+l ij 1 s=0 j J t T x js 1 x jt {0, 1} j J (i, j) E, t T j J, t T Totally unimodular matrix, integer polyhedron min w j (t)x jt x jt = 1 t T j J t T T T tx js tx is l ij t=0 t=0 x jt {0, 1} Polyhedron is not integer! j J (i, j) E j J, t T For the RCPSP, the LP relaxation of the time-indexed model with disagreggated precedence constraints is tighter. Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 25 / 78

72 MILP Formulations with time-indexed variables : facets Facet-inducing inequalities for the single-machine polyhedron (without precedence constraints) [SW92] x jt 1 j J t T t x js 1 t T j J s=t p j +1 t+ 1 x js + t x is 1 s=t p j +1 i j s=t p i + j J, t T, All facet-inducing inequalities with rhs = 1 [vhs99]. {2,..., max p i } i j Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 26 / 78

73 Sousa and Wolsey valid inequalities [SW92] T u=1 t i J s=t p i +1 t+ 1 i J s=t p i + x ju 2 = 1 2 x is x is T u=1 x ju 2 + n t i=1 s=t p i +1 t+ 1 x is 2 + i J s=t p j +1 x js + i j t+ 1 s=t p i + t s=t p i + x is x is 1 Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 27 / 78

74 Sousa and Wolsey valid inequalities [SW92] For each time period t T, for each task j T and for each {2,..., max i j p i } t p j + 1 t + 1 j t p i + t 1 i J \ {j} Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 28 / 78

75 van den Akker et al. valid ineqalities [vhs99] j 1 u 1 p j1 + 1 u u 2 p j1 v u z u 2 p j1 + 1 u u 1 p j2 + 1 u1 max{u 2 v, u } p j2 min{u z, u 2 } u 2 p j2 + 2 u2 j 2 2 u 1 p i z + 1 u 1 u 2 p i + 1u u 2 p i + 2 u 2 v 1 i J\{j 1, j 2 } L M U With the same principle all facet-inducing inequalities with rhs = 2 are derived. In [WJNS02], relation with valid ineqalities for the node packing problem is established. Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 29 / 78

76 Valid inequalities for the time-indexed RCPSP ILP Hardin et al. [HNS08] extend the Sousa and Wolsey 1-machine cuts to the RCPSP. Let F be a forbidden set (or cover), i.e. j F b i > B. j F ts=t pj +1 x js F 1 is a valid inequality t T. Consider now a special task j F and an interval v 0 ts=t pj i F \{j} +1+v x iq + t+v s=t p j +1 x js F 1 is a valid inequality t T. The inequality defines a facet for the polyhedron reduced to jobs in C if C is a minimal forbidden set, i.e. C\{i} b i B, i C. Hardin et al. propose lifting procedures and conditions for the resulting inequalities to be facet-inducing for the complete polyhedron. Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 30 / 78

77 Valid inequalities for the time-indexed RCPSP ILP Hardin et al. [HNS08] extend the Sousa and Wolsey 1-machine cuts to the RCPSP. Let F be a forbidden set (or cover), i.e. j F b i > B. j F ts=t pj +1 x js F 1 is a valid inequality t T. Consider now a special task j F and an interval v 0 ts=t pj i F \{j} +1+v x iq + t+v s=t p j +1 x js F 1 is a valid inequality t T. The inequality defines a facet for the polyhedron reduced to jobs in C if C is a minimal forbidden set, i.e. C\{i} b i B, i C. Hardin et al. propose lifting procedures and conditions for the resulting inequalities to be facet-inducing for the complete polyhedron. Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 30 / 78

78 Valid inequalities for the time-indexed RCPSP ILP Hardin et al. [HNS08] extend the Sousa and Wolsey 1-machine cuts to the RCPSP. Let F be a forbidden set (or cover), i.e. j F b i > B. j F ts=t pj +1 x js F 1 is a valid inequality t T. Consider now a special task j F and an interval v 0 ts=t pj i F \{j} +1+v x iq + t+v s=t p j +1 x js F 1 is a valid inequality t T. The inequality defines a facet for the polyhedron reduced to jobs in C if C is a minimal forbidden set, i.e. C\{i} b i B, i C. Hardin et al. propose lifting procedures and conditions for the resulting inequalities to be facet-inducing for the complete polyhedron. Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 30 / 78

79 Valid inequalities for the time-indexed RCPSP ILP Hardin et al. [HNS08] extend the Sousa and Wolsey 1-machine cuts to the RCPSP. Let F be a forbidden set (or cover), i.e. j F b i > B. j F ts=t pj +1 x js F 1 is a valid inequality t T. Consider now a special task j F and an interval v 0 ts=t pj i F \{j} +1+v x iq + t+v s=t p j +1 x js F 1 is a valid inequality t T. The inequality defines a facet for the polyhedron reduced to jobs in C if C is a minimal forbidden set, i.e. C\{i} b i B, i C. Hardin et al. propose lifting procedures and conditions for the resulting inequalities to be facet-inducing for the complete polyhedron. Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 30 / 78

80 Valid inequalities for the time-indexed RCPSP ILP Hardin et al. [HNS08] extend the Sousa and Wolsey 1-machine cuts to the RCPSP. Let F be a forbidden set (or cover), i.e. j F b i > B. j F ts=t pj +1 x js F 1 is a valid inequality t T. Consider now a special task j F and an interval v 0 ts=t pj i F \{j} +1+v x iq + t+v s=t p j +1 x js F 1 is a valid inequality t T. The inequality defines a facet for the polyhedron reduced to jobs in C if C is a minimal forbidden set, i.e. C\{i} b i B, i C. Hardin et al. propose lifting procedures and conditions for the resulting inequalities to be facet-inducing for the complete polyhedron. Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 30 / 78

81 Valid inequalities for the time-indexed RCPSP ILP Hardin et al. [HNS08] extend the Sousa and Wolsey 1-machine cuts to the RCPSP. Let F be a forbidden set (or cover), i.e. j F b i > B. j F ts=t pj +1 x js F 1 is a valid inequality t T. Consider now a special task j F and an interval v 0 ts=t pj i F \{j} +1+v x iq + t+v s=t p j +1 x js F 1 is a valid inequality t T. The inequality defines a facet for the polyhedron reduced to jobs in C if C is a minimal forbidden set, i.e. C\{i} b i B, i C. Hardin et al. propose lifting procedures and conditions for the resulting inequalities to be facet-inducing for the complete polyhedron. Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 30 / 78

82 Outline 1 Introduction 2 Polyhedral studies and cutting plane generation 3 (Mixed) integer programming for scheduling problems Time-indexed variables Total unimodularity Sequencing and natural date variables Flow (TSP) and natural date variables Positional date and assignment variables Hybrid formulations 4 Column generation 5 A few Computational results Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 31 / 78

83 Sequencing variables Time-indexed models yield good LP relaxation but the number of variables can be huge = need to consider more compact models. Binary variable y ij = 1 if and only if S j S i + p i Disjunctive case : y ij = 1 y ji, (i, j) D (half of the variables can be dropped) and y ij is the incidence vector of linear orderings : If there are no release dates and in the 1-machine case C j = i D\{j} y ij p i + p j Not possible to consider start dependent costs = objective i J w i C i Christian Artigues Scheduling and (Integer) Linear Programming CPAIOR 2012, Nantes 32 / 78