This paper presents a combination of variable neighbourhood search and mathematical programming to minimize the sum of earliness and tardiness penalty costs of all operations for just-in-time job-shop scheduling problem (JITJSSP). Unlike classical E/T scheduling problem with each job having its earliness or tardiness penalty cost, each operation in this paper has its earliness and tardiness penalties, which are paid if the operation is completed before or after its due date. Our hybrid algorithm combines (i) a variable neighbourhood search procedure to explore the huge feasible solution spaces efficiently by alternating the swap and insertion neighbourhood structures and (ii) a mathematical programming model to optimize the completion times of the operations for a given solution in each iteration procedure. Additionally, a threshold accepting mechanism is proposed to diversify the local search of variable neighbourhood search. Computational results on the 72 benchmark instances show that our algorithm can obtain the best known solution for 40 problems, and the best known solutions for 33 problems are updated.
1. Introduction
The importance of “just-in-time” inventory management in industry has motivated the study of theoretical scheduling models. Among these models, many research efforts have been devoted to the classical earliness and tardiness (E/T) scheduling problems, and most of them consider earliness and tardiness penalties only for the last operation of the job [1–3]. Although the job may be finished on time, no storage and insurance cost are considered for intermediary operations which are processed as early as possible, which obviously violate the JIT philosophy [4].
For the just-in-time job-shop scheduling problem (JITJSSP), a job is decomposed in a sequence of operations that should be performed in a specific order. Each operation has its due date and earliness-tardiness penalties, which are paid if the operation is completed before or after its due date. Thus, the unwanted intermediate storage costs are avoided by finishing the operation as close to its due date as possible [4].
Baptiste et al. [4] defined an integer programming model for the JITJSSP problem and proposed methods based on two Lagrangian relaxations of the model to derive lower and upper bounds. They reported good lower and upper bounds for 72 instances. However, the gaps between lower and upper bound are still large for most instances. Monette et al. [5] later presented a constraint programming approach, which includes a novel filtering algorithm and dedicated heuristics. Although the method in [6] can give an inserted idle time schedule solution at last, it cannot offer the optimum (even near-optimum) solution for the JITJSSP with loose due date environment because few improvements occur on the instances of small scale.
In order to minimize the sum of total penalty costs in the JITJSSP problem, all operations are required to be completed as close to their due dates as possible, and thus the idle time is necessary to delay the start time of an early operation. In fact, it is a serious trouble how well idle time is inserted when severity resource contention occurs [7]. The mathematical programming method, which is suitable for a small solution space, can obtain an optimal solution. However, for the JITJSSP problem with a large number of jobs and machines, the computation time is intractability.
As for the JITJSSP problem with nonregular performance measure, how to explore the huge feasible space and how to insert machine idle time are two key issues. To deal with these two issues in solving the JITJSSP problem more effectively and efficiently, the hybrid algorithm is proposed in this paper with the mathematical programming (MP) model embedded to each iteration procedure of variable neighborhood search (VNS). In each iteration procedure of VNS, a sequence of operations on machines is produced by exploring the two swap and insertion neighborhoods of the current solution, and the mathematical programming model is then employed to insert idle time and optimize the completion times of the operations for the given sequences.
The rest of the paper is organized as follows. Section 2 discusses the problem formulation. The overall operational structure of the VNS/MP algorithm is presented in Section 3. An extensive variable neighborhood search study is discussed in Section 4. The mathematical programming model is presented in Section 5. A benchmarking study and experimental results are discussed in Section 6. Finally, the conclusions are given in Section 7.
2. Problem Formulation
The following definition of the JITJSSP problem is given on Baptiste et al. [4]. Consider a finite set of n jobs J={J1,J2,…,Jn} that have to be processed on a finite set of m machines M={M1,M2,…,Mm}. Each job Ji has a sequence of m operations, Oi={oi1,oi2,…,oim}, that need to be processed in sequence. The kth operation of job i, oik, has to be processed by a prescribed machine M(oik)∈M. Likewise, the set of operations that must be processed on each machine, Mu∈M, is denoted by O(Mu). Without loss of generality, it is assumed that at most one operation of each job is processed on each machine, each machine can process at most one operation at a time, and no preemption is allowed. The notations used in the problem formulation are as follows:
pik: processing time (duration) of operation oik;
dik: due date for operation oik, assuming dik+pik+1≤dik+1;
stik: start time of operation oik;
cik: completion time of operation oik;
Eik: earliness of operation oik, calculated by max(0,dik-cik);
Tik: tardiness of operation oik, calculated by max(0,cik-dik);
αik: earliness penalty per unit time for operation oik;
βik: tardiness penalty per unit time for operation oik.
Object Function. We have
(1)Min ∑i=1n∑k=1m(αikEik+βikTik)iiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiii(2)Subjecttocik+1≥cik+pik+1,i=1,2,…,n∧k=1,2,…,m-1(3)cik≥cjl+pik∨cjl≥cik+pjl,iii∀oik,ojl∈O(Mu)∧i≠j(4)sti1≥0,i=1,2,…,n.
The objective function (1) is to minimize the sum of earliness and tardiness costs of all operations. For each pair of consecutive operations oik and oik+1 of the same job i, the precedence constraint (2) ensures that operation oik+1 cannot be processed before operation oik is completed. The resource constraint (3) ensures that two operations of two different jobs i and j, oik and ojl, in the set O(Mu), cannot be processed simultaneously on machine Mu. Equation (4) ensures that the first operation of each is released after time zero.
3. The Overall Operational Structure of the VNS/MP Algorithm
Variable neighborhood search [8, 9] is a new local search technique that tries to escape from local optimum by changing neighborhood structures. In its basic form, VNS explores increasingly distant neighborhoods of the current incumbent solution, makes a local search from a neighbor solution to a local optimum, and jumps to it if and only if an improvement has been made.
The overall operational structure of the VNS/MP algorithm for the JITJSSP problem is shown in Figure 1. The VNS method is employed to explore the huge feasible space and provide a feasible schedule solution and, in initialization, shaking, and local search procedure of VNS. The mathematical programming model is introduced to make the operations complete closer to its due dates by inserting machine idle times into the given sequences and then feed backs an optimal schedule solution to VNS for its next iterative optimization procedure.
Overall operational structure of the VNS/MP algorithm.
The overall steps of the VNS method are shown in Figure 1. Let Nk(k=0,1,…,kmax-1) denote neighbourhood structures and Nk(x) the set of the solutions in the kth neighbourhood of x. To avoid consuming too much computational time, the maximum number of neighbourhood structures, kmax, is suggested to be 2 [10]. The two kinds of neighbourhood structures, swap and insertion, are designed to change and improve the schedule solution for the shaking and local search procedure. Let N0 indicate the neighbourhood swap and let N1 indicate the neighbourhood Insertion. The stopping condition is maximum number of iterations N.
4. Variable Neighbourhood Search: Explore the Solution Space
To achieve greater efficiency in VNS, we propose an excellent initial solution generation method and two kinds of neighborhood structures, swap and insertion, based on the characteristic of JITJSSP problem. Moreover, the local search procedure with the filter mechanism is designed to eliminate the swap movement that is unlikely to lead to good schedules. The key three issues in VNS, initialization, neighbourhood structures, and local search, are described in detail below.
4.1. Initialization
A good initial solution is able to reduce the computational time to achieve the optimal solution. Therefore, we present an initial solution generator that involves two new dispatching rules, each producing a candidate solution and the better one being selected as an initial solution. The rule, as a variant of the Earliest Due Date (EDD) rule, is that the operation with the earliest due date is selected first.
The other rule is a modification to ALL + CR + SPT [11]. Two separate sequences are produced for each machine by this new rule and the best solution is selected. The overdue jobs are placed in the priority sequence, in which the machine chooses the job whose pik/βik is minimal to be processed. But if this sequence is empty, the machine will choose the job whose pik/αik is maximal to be processed in the other sequence. For the criteria mean absolute lateness, ALL + CR + SPT can achieve the best performance at certain tightness levels. The new rule inherits this quality and can make the operation completion time closer to its due date.
4.2. Neighbourhood Structures: Swap and Insertion
Firstly, the immediate job predecessor and the immediate job successor of the operation u are denoted as JP[u] and JS[u], respectively. Similarly, the immediate machine predecessor and the immediate machine successor of u are denoted as MP[u] and MS[u], respectively. Besides, two successive operations u and v on the same machine (providing u is processed before v) are adjacent when there is no idle time between them (i.e., cu=stv). And then u is the adjacent machine predecessor of v while v is the adjacent machine successor of u. Finally, a block is defined as a maximal sequence of size at least two, consisting of adjacent operations that are processed on the same machine.
The two neighbourhood structures, swap and insertion, are applied in shaking and local search procedures of VNS are described as follows.
4.2.1. Swap Neighbourhood
Swap the selected operation with its adjacent machine successor while it is early or when its adjacent machine predecessor is tardy.
In each iteration of VNS, the schedule solution obtained from mathematical programming model is the optimal solution for the current sequence, and the completion time of each operation is as close to its due date as possible. The first step of generating a neighbour in both neighbourhoods is to randomly select an operation which is either early and has an adjacent machine successor or tardy and has an adjacent machine predecessor. The selected operation by this Swap procedure has the two following lemmas.
Lemma 1.
If the selected operation u is early and has no adjacent machine successor (i.e., the immediate machine successor does not exist or there is idle time between u and the immediate machine successor), the immediate job successor of u is also early.
Proof.
If the selected operation u is early and has no adjacent machine successor, thus, cu<du, where cu and du denote the completion time and due date of u, respectively. cu is equal to the start time of its immediate job successor; that is, cu=stJS[u]. There is cJS[u]=stJS[u]+pJS[u]=cu+pJS[u]<du+pJS[u]≤dJS[u], which means that the immediate job successor is also early.
Lemma 2.
If the selected operation u is tardy and has no adjacent machine predecessor (i.e., the immediate machine predecessor does not exist or there is idle time between u and the immediate machine predecessor), the immediate job predecessor of u is also tardy.
Proof.
If the selected operation u is tardy and has no adjacent machine predecessor, the start time of u must be equal to the completion time of the immediate job predecessor; that is, stu=cJP[u]. Thus, cJP[u]=stu>du-pu≥dJP[u], which means that the immediate job predecessor is also tardy.
If the selected operation is early and has no adjacent machine successor, we will trace its job successors until finding one job successor which has an adjacent machine successor. If the job successor is early according to Lemma 1, then the job successor will be chosen to move. This Swap procedure is shown in Figure 2.
Example for swap procedure.
In Figure 2, let u be the selected operation; it is early and there is idle time between u and its immediate machine successor v. We trace the job successors of u, and the operation JS[u] meets exactly the swap condition. Thus, the operation JS[u] is to be swapped with the operation JP[v] to generate a new neighbour with good performance.
4.2.2. Insertion Neighbourhood
For block of a solution, it is not only the result of operations gathering together to compete for one machine but also an obstacle which prevents the operation from closing to its due date. Therefore, inserting the selected operation into the start or the end of the block can exploit the idle times effectively between the blocks for a performance improvement.
Let the selected operation be u, let the block which u belongs to be B, and let v (or w) denote the first (or last) operation of B. The insertion procedure varies with the status of u as follows.
Insertion Procedure When u Is Early. When the operation w is the last one of block B, move u to the right of w to bring new neighbour if stw≤stJS[u]; otherwise, select an operation x randomly which is processed after u and move u to the right of w to bring new neighbour if stx≤stJS[u]. See Figure 3.
Insertion procedure for early operation.
Insertion Procedure When u Is Tardy. When the operation v is the last one of block B, move u to the left of v to bring new neighbour if cv≥cJP[u]; otherwise, select an operation x randomly which is processed before u and move u to the left of v to bring new neighbour if cx≥cJP[u]. See Figure 4.
Insertion procedure for tardy operation.
4.3. Local Search with the Filter Mechanism
As for the local search procedure in VNS, we make two improvements: threshold accepting mechanism to diversify the search space and filter mechanism to avoid the possible performance deterioration of schedule solutions in swap procedure.
4.3.1. Local Search with Threshold Accepting Mechanism
The threshold accepting mechanism [12] is embedded into each iteration process of local search procedure [13]. The current solution x′ is replaced by the new solution x′′ when the performance deviation between x′ and x′′ is less than a threshold dr. Set dr=f(x′)/2 in this paper. Furthermore, in order to enhance the search efficiency, a certain number s candidates are selected randomly from the set Nk(x′), and then let the best one in s candidates replace x′′. See Algorithm 1.
Algorithm 1: Overall operational structure of local search.
(1) Get initial solution x′, dr, and s.
(2) Set q←0 and u←0.
(3) While q<qmax do
(a) x′′← the best one among the s candidates from Nu(x′).
(b) If f(x′′)-f(x′)≤dr then x′←x′′;
else u←|u-1|.
(c) q←q+1.
(4) x′′←x′.
4.3.2. Filter Mechanism for Swap Procedure
For a given sequence produced by VNS, an optimal schedule solution can be obtained by mathematical programming model. Thus, swapping two adjacent operations may delay the completion time of the successors and lead the new neighbour quality to deteriorate. In order to solve this problem, a filter mechanism is proposed in local search to prevent a new neighbour by swap operation from deteriorating the performance of schedule solutions.
Providing the adjacent operations to be swapped, u and v, we introduce overlap to evaluate effect of each swap operation. Figures 5(a) and 5(b) show two kinds of overlap.
Overlap degree of two operations and its impact on swap procedure.
In Figure 5(a), there is a large overlap between two operations, JP[v] and u, as well as JS[u] and v, before swap operation. However, the completion times of u and JS[u] are more seriously delayed once the swap operation is executed. On the other hand, in Figure 5(b), although there is also an overlap between two operations, JP[v] and u, the swap operation can improve the solution because of the machine idle time.
In order to improve the solution and avoid the deterioration in swap operation, the overlap degree of two operations u and v is evaluated by (5), and the swap procedure can be executed if Lap(u,v)≤Lapmax. Lapmax refers to a maximum of overlap degree:
(5)Lap(u,v)=max(cJP[v],cMP[u])-stu+cv-min(stJS[u],stMS[v])pu+pv.
5. Mathematical Programming Model: Optimize the Completion Times of the Operations for a Given Solution
For the VNS/MP algorithm, three steps in VNS procedure can produce three feasible solutions, respectively.
Initialization step: an initial feasible solution is produced by EDD or ALL-JIT rule.
Shaking step: a feasible solution is generated by insertion or swap operations.
Local search step: a new neighbor and a given solution are also produced by insertion or swap operations.
All these three feasible solutions have satisfied the resource constraint (3), which can ensure that only one operation is processed simultaneously on one machine. Thus, by VNS procedure, the difficult resource constraint (3) is relaxed as follows:(6)stjl+pjl≤stik,M(oik)=M(ojl),
where oik is processed after ojl.
Therefore, in our proposed VNS/MP algorithm, VNS is used to explore the huge search space and achieve a good feasible solution. The resource constraint in JIT scheduling model is relaxed to make a good chance for the improvement by mathematical programming method. Thus, the JIT scheduling model can be changed into mathematical programming model, and the optimal completion times of all the operations for a given solution can be achieved in a short computation time.
6. Computational Results and Discussion6.1. JITJSSP Instances
We used the same set of 72 instances generated by Baptiste et al. [4]. Each instance is named I-n-m-DD-W-ID, where I refers to “instance” and n and m are the number of jobs and machines, respectively, which have 9 combinations totally; that is, (n×m) ∈ ({10, 15, 20} × {2, 5, 10}). Each combination of n and m contains 8 instances such as DD, W, and ID which are chosen randomly.
DD denotes the distances between due dates of consecutive operations which are exactly equal to the processing time of the last operation (DD = tight) or equal to the processing time plus a random value in [0,10] (DD = loose).
W denotes the relation between earliness and tardiness penalty. There is W = equal if both α and β are chosen randomly in [0.1,1] or W = tardy if α is chosen randomly in [0.1,0.3] and β in [0.1,1].
ID is set to be 1 or 2 to identify the two instances generated for each combination of the other parameters.
The VNS method is executed by a fixed number of N=100 iterations. The maximum number of local search iterations is qmax=50. Set the number of the candidates and the upper bound of overlap to be s=3 and Lapmax=0.5, respectively.
6.2. Results and Comparison
The VNS/MP algorithm is coded in Matlab and the mathematical programming model is implemented by the Xpress optimizer in FICO Xpress.
The solution values for instances with n=10, 15, and 20 jobs are reported, respectively, on Tables 1, 2, and 3. Table 4 shows the total number of best solutions for the 72 instances. The columns LB and UB list the best of the lower bounds and the upper bound reported by Baptiste et al. [4]. The column CP refers to the constraint programming approach proposed by Monette et al. [5]. The column EA/LS/MP refers to the combination of evolutionary algorithm, local search, and mathematical programming presented by Dos Santos et al. [6]. The best result of each instance computed by these five methods is highlighted in boldface type.
Results for instances with 10 jobs.
Instance
LB [4]
UB [4]
CP [5]
EA/LS/MP [6]
VNS/MP
I-10-2-tight-equal-1
434
453
461.96
463.58
461.96
I-10-2-tight-equal-2
418
458
448.32
448.32
448.32
I-10-5-tight-equal-1
660
826
764.80
765.35
689.11
I-10-5-tight-equal-2
612
848
779.40
779.52
763.24
I-10-10-tight-equal-1
1126
1439
1339.64
1281.66
1277.44
I-10-10-tight-equal-2
1535
2006
1902.30
2175.22
1878.26
I-10-2-loose-equal-1
219
225
224.84
224.84
224.84
I-10-2-loose-equal-2
313
324
319.37
331.08
319.37
I-10-5-loose-equal-1
1263
1905
1823.85
1740.08
1774.39
I-10-5-loose-equal-2
878
1010
999.14
1045.76
967.73
I-10-10-loose-equal-1
331
376
381.88
763.84
364.39
I-10-10-loose-equal-2
246
260
256.78
343.12
249.85
I-10-2-tight-tard-1
174
195
179.46
179.68
179.46
I-10-2-tight-tard-2
143
147
164.38
146.43
145.37
I-10-5-tight-tard-1
361
405
398.37
410.16
387.30
I-10-5-tight-tard-2
461
708
639.16
632.91
635.93
I-10-10-tight-tard-1
574
855
773.26
777.25
687.65
I-10-10-tight-tard-2
666
800
830.39
1031.13
779.30
I-10-2-loose-tard-1
416
416
416.44
416.44
416.44
I-10-2-loose-tard-2
137
138
137.94
143.00
137.94
I-10-5-loose-tard-1
168
188
182.64
259.02
175.08
I-10-5-loose-tard-2
355
572
513.91
509.33
499.93
I-10-10-loose-tard-1
356
409
387.05
438.04
383.86
I-10-10-loose-tard-2
138
152
144.94
169.07
144.94
Sum of best solutions
2
6
4
20
Results for instances with 15 jobs.
Instance
LB [4]
UB [4]
CP [5]
EA/LS/MP [6]
VNS/MP
I-15-2-tight-equal-1
3316
3559
3641.19
3344.54
3344.54
I-15-2-tight-equal-2
1449
1579
1534.12
1480.97
1479.76
I-15-5-tight-equal-1
1052
1663
1504.04
1369.78
1402.99
I-15-5-tight-equal-2
1992
2989
2993.50
2728.02
2693.24
I-15-10-tight-equal-1
4389
8381
8189.70
7340.27
7395.36
I-15-10-tight-equal-2
3539
7039
5536.07
5399.20
6018.21
I-15-2-loose-equal-1
1032
1142
1249.68
1041.70
1041.70
I-15-2-loose-equal-2
490
520
524.10
506.44
497.97
I-15-5-loose-equal-1
2763
4408
3745.96
3267.79
3302.61
I-15-5-loose-equal-2
2818
4023
3397.42
3441.69
3760.00
I-15-10-loose-equal-1
758
1109
1033.06
1078.93
986.43
I-15-10-loose-equal-2
1242
2256
1792.67
1996.26
1563.03
I-15-2-tight-tard-1
786
913
835.52
790.61
790.50
I-15-2-tight-tard-2
886
956
947.17
905.73
905.37
I-15-5-tight-tard-1
1014
1538
1530.96
1389.81
1405.01
I-15-5-tight-tard-2
626
843
775.01
712.50
732.15
I-15-10-tight-tard-1
649
972
921.67
926.22
813.46
I-15-10-tight-tard-2
955
1656
1663.05
1304.27
1459.33
I-15-2-loose-tard-1
650
730
666.37
655.35
654.84
I-15-2-loose-tard-2
278
310
336.48
291.68
293.17
I-15-5-loose-tard-1
1098
1723
1478.97
1315.53
1396.54
I-15-5-loose-tard-2
314
374
401.65
396.14
396.06
I-15-10-loose-tard-1
258
312
300.11
356.26
282.35
I-15-10-loose-tard-2
476
855
658.90
710.55
778.40
Sum of best solutions
1
1
12
12
Results for instances with 20 jobs.
Instance
LB [4]
UB [4]
CP [5]
EA/LS/MP [6]
VNS/MP
I-20-2-tight-equal-1
1901
2008
2115.58
1940.30
1941.18
I-20-2-tight-equal-2
912
1014
1104.20
948.86
943.70
I-20-5-tight-equal-1
2506
3090
3349.28
2933.28
2853.31
I-20-5-tight-equal-2
5817
7537
7883.50
6915.06
7038.92
I-20-10-tight-equal-1
6708
12951
14004.90
10520.40
11663.30
I-20-10-tight-equal-2
5705
9435
8535.88
7451.68
7201.05
I-20-2-loose-equal-1
2546
2708
2789.07
2551.60
2550.53
I-20-2-loose-equal-2
3013
3318
3386.88
3109.29
3213.52
I-20-5-loose-equal-1
6697
9697
9481.56
7646.90
8231.23
I-20-5-loose-equal-2
6017
8152
8835.72
7294.50
7642.51
I-20-10-loose-equal-1
3538
6732
6101.67
5022.49
5753.52
I-20-10-loose-equal-2
1344
2516
1963.05
1984.47
1816.53
I-20-2-tight-tard-1
1515
1913
1892.22
1682.72
1685.84
I-20-2-tight-tard-2
1375
1594
1704.26
1457.13
1452.05
I-20-5-tight-tard-1
3244
4147
4067.73
3640.00
3990.04
I-20-5-tight-tard-2
1633
1916
2040.70
1873.80
1911.91
I-20-10-tight-tard-1
3003
5968
5125.88
4778.16
5290.52
I-20-10-tight-tard-2
2740
3788
3938.51
3270.09
3491.52
I-20-2-loose-tard-1
1194
1271
1409.73
1206.97
1204.92
I-20-2-loose-tard-2
735
857
907.60
774.22
788.36
I-20-5-loose-tard-1
2524
3377
4015.62
2973.23
3720.83
I-20-5-loose-tard-2
3060
5014
4539.36
3654.86
3719.80
I-20-10-loose-tard-1
2462
6237
7287.00
5100.46
6169.68
I-20-10-loose-tard-2
1226
1830
1727.88
1634.58
1588.70
Sum of best solutions
0
0
16
8
Total number of best solutions for the 72 instances.
Algorithm
UB [4]
CP [5]
EA/LS/MP [6]
VNS/MP
Total number of best solutions
3
7
32
40
From Table 4, our VNS/MP algorithm obtains the best solution on 40 of the 72 instances and outperforms all of its competitors. Although VNS/MP has overwhelming dominance over the UB and CP, it is a little better than EA/LS/MP. Compared with EA/LS/MP, VNS/MP gives better solutions for most of the small instances; however, as the size of instances increases its performance deteriorates increasingly.
Yet we should not make a conclusion that VNS/MP is not suitable for solving JITJSSP problems of large scale. The neighborhood structures swap and insertion in VNS/MP are based on adjacent operations, which result in the solution space increasing along with the number of adjacent operations to be operated. As a consequence, the search space will become too huge when the due date setting is tight which makes most of the operations tend to be adjacent. Practically, most of the due dates of the 72 benchmark instances are extremely tight, which exactly coincides with the advantage of EA/LS/MP. In other words, VNS/MP solves the JIT scheduling problems with loose due dates effectively but its performance may decline when it is used to schedule problems with tight due dates.
7. Conclusions
We propose and evaluate a VNS/MP hybrid algorithm to the JITJSSP problem that searches a minimal sum of earliness and tardiness penalty costs for all operations, not only for the jobs. This hybrid method combines (i) the variable neighbourhood search procedure to explore the huge feasible space of the JITJSSP problem efficiently by escaping from local optima with alternating the swap and insertion neighborhood structures and (ii) the mathematical programming model to insert idle machine time and optimize the completion times of the operations for a given sequence.
The VNS/MP hybrid algorithm is tested with 72 benchmark instances in JITJSSP problem and compared with three other approaches. It can find the best known solution for 40 instances with minimal computational effort, and the best known solutions for 33 instances are updated. As continuation to this work, we intend to design more efficiency neighborhood structures to improve the performance of VNS method to meet the large-scale JITJSSP problems with tight due date environment.
Conflict of Interests
The authors declare that there is no conflict of interests regarding the publication of this paper.
BakerK. R.ScudderG. D.Sequencing with earliness and tardiness penalties: a review1990381223610.1287/opre.38.1.22MR1041230ZBL0699.90052ZhuZ.HeadyR. B.Minimizing the sum of earliness/tardiness in multi-machine scheduling: a mixed integer programming approach20003822973052-s2.0-003422808410.1016/S0360-8352(00)00048-6ThiagarajanS.RajendranC.Scheduling in dynamic assembly job-shops to minimize the sum of weighted earliness, weighted tardiness and weighted flowtime of jobs20054944635032-s2.0-2804445269110.1016/j.cie.2005.06.005BaptisteP.FlaminiM.SourdF.Lagrangian bounds for just-in-time job-shop scheduling200835390691510.1016/j.cor.2006.05.009MR2566397ZBL1278.90142MonetteJ.-N.DevilleY.Van HentenryckP.Just-in-time scheduling with constraint programmingProceedings of the 19th International Conference on Automated Planning and Scheduling (ICAPS '09)September 20092412482-s2.0-78650619326Dos SantosA. G.AraujoR. P.ArroyoJ. E. C.A combination of evolutionary algorithm, mathematical programming, and a new local search procedure for the just-in-time job-shop scheduling problem2010607310242-s2.0-7795460179010.1007/978-3-642-13800-3_2KanetJ. J.SridharanV.Scheduling with inserted idle time: problem taxonomy and literature review2000481991102-s2.0-0033685372HansenP.MladenovićN.Variable neighborhood search: principles and applications2001130344946710.1016/S0377-2217(00)00100-4MR1816664ZBL0981.90063ChengW.GuoP.ZhangZ.ZengM.LiangJ.Variable neighborhood search for parallel machines scheduling problem with step deteriorating jobs2012201220928312MR2971182ZBL1264.90086LiaoC.-J.ChengC.-C.A variable neighborhood search for minimizing single machine weighted earliness and tardiness with common due date20075244044132-s2.0-3424817901910.1016/j.cie.2007.01.004ChengT. C. E.JiangJ.Job shop scheduling for missed due-date performance1998342-42973072-s2.0-0032041372BouffardV.FerlandJ. A.Improving simulated annealing with variable neighborhood search to solve the resource-constrained scheduling problem200710637538610.1007/s10951-007-0043-7MR2373975ZBL1153.90415ZandiehM.AdibiM. A.Dynamic job shop scheduling using variable neighbourhood search2010488244924582-s2.0-7795111971010.1080/00207540802662896