Hindawi

Advances in Materials Science and Engineering

Volume 2021, Article ID 3732720, 15 pages
https://doi.org/10.1155/2021/3732720

Research Article
A Heuristic Approach for Two-Dimensional Rectangular Cutting
Stock Problem considering Balance for Material Utilization and
Cutting Complexity

1,2 1,2
Dianjian Wu and Guangyou Yang
1
Agricultural Machinery Engineering Research and Design College, Hubei University of Technology, Wuhan 430068, China
2
Hubei Province Agricultural Equipment Intelligent Engineering Technology Center, Wuhan 430068, China

Correspondence should be addressed to Guangyou Yang; [email protected]

Received 12 May 2021; Accepted 10 October 2021; Published 8 November 2021

Academic Editor: Francisco Javier Fernández Fernández

Copyright © 2021 Dianjian Wu and Guangyou Yang. -is is an open access article distributed under the Creative Commons
Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is
properly cited.
-e common staged patterns are always required during the cutting process for separating a set of rectangular items from
rectangular plates in manufacturing industries. Two-staged patterns can reduce cutting complexity at the expense of material
utilization; three-staged patterns do the opposite. Combining these two types of staged patterns may be a good balance for two
contradictory objectives of material utilization and cutting complexity. A heuristic approach is proposed to solve the two-di-
mensional rectangular cutting stock problem with a combination of two-staged general patterns (2SGP) and three-staged
homogenous patterns (3SHP). Firstly, the 2SGP and 3SHP are constructed by using recursive techniques. -e pattern with the
larger value is selected as the candidate pattern. -en, the value of each item is corrected according to the current candidate
pattern. A cutting plan accurately satisfying all items demand is obtained by using the sequential heuristic algorithm. Finally, the
cutting plan with a minimum number of used plates is achieved by applying the iterative algorithm. -e computational results
indicate that the proposed heuristic approach is more effective in material utilization and cutting complexity than the two
published algorithms with staged patterns.

1. Introduction common application restrictions of “Guillotine cuts” and

“Orientation.” Based on these two restrictions, the
-e two-dimensional rectangular cutting stock problem 2DRSSSCSP can be categorized as four variants, including
(2DRCSP) is always encountered in many manufacturing 2DRSSSCSP-GO, 2DRSSSCSP-GR, 2DRSSSCSP-FO, and
industries, such as steel products, paper, wood, and glass. A set 2DRSSSCSP-FR, which are shown in Figure 1. G represents
of rectangular items with specific demand need to be produced that the cutting style must be guillotine cuts, where each cut
from cutting the same rectangular plates with the optimization must be performed from edge-to-edge parallel to each side of
objective of minimizing the number of used plates. -is feature the used rectangular plate. F represents that the cutting style
of the 2DRCSP is referred to as a two-dimensional rectangular can be free, namely, not considered guillotine cuts. O de-
single stock size cutting stock problem (2DRSSSCSP) in notes that the items are oriented, that is, not rotated. R
Wäscher et al. [1]. -e solution of the 2DRSSSCSP is a cutting denotes that the items can be rotated by 90°.
plan including a set of different patterns with their frequencies. -is paper mainly focuses on the research of the
-erefore, how to design patterns with optimization strategy 2DRSSSCSP-GR, where the cutting style must be guillotine
becomes the most noteworthy problem. cuts, and each item is allowed to be rotated by 90°. -e staged
Some practical restrictions need to be observed while patterns are always constructed to satisfy guillotine cuts. -e
solving the 2DRSSSCSP. Lodi et al. [2] introduced two common staged patterns are two-staged patterns and three-
 5928, 2021, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2021/3732720 by Readcube (Labtiva Inc.), Wiley Online Library on [24/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
2 Advances in Materials Science and Engineering

staged patterns. In general, the two-staged patterns have 3SHP are built by using recursive techniques. -e candidate
simpler cutting complexity but less material utilization than pattern with the objective of a large pattern value can be
the three-staged patterns. Two-staged patterns are welcomed chosen from the generated 2SGP and 3SHP. -en, the value
to reduce cutting complexity when the material cost is of each item is corrected according to the current candidate
relatively cheaper than the guillotine cuts cost. -ree-staged pattern. A cutting plan including 2SGP and 3SHP is achieved
patterns are selected to improve material utilization oth- with a sequential heuristic algorithm until all items demands
erwise. It is difficult to consider two contradictory objectives are fulfilled. Finally, the cutting plan with the minimum
of material utilization and cutting complexity during solving number of used plates is achieved by applying the iterative
the 2DRSSSCSP-GR with only two-staged patterns or three- algorithm.
staged patterns. -erefore, combining two types of staged -e remainder of the paper is organized as follows: the
patterns for the 2DRSSSCSP-GR may be an effective ap- related literature is reviewed in Section 2. -e generation
proach to balance material utilization and cutting procedures for the 2SGP and 3SHP are introduced in Section 3.
complexity. An iterative sequential heuristic procedure (ISHP) considering
We firstly introduce two-staged patterns and three- both 2SGP and 3SHP is proposed for solving the 2DRSSSCSP-
staged patterns based on two types of strips in Figure 2. Each GR in Section 4. -e comparison results are reported in Section
homogenous strip in Figure 2(a) is only composed of 5, followed by the conclusions in Section 6.
multiple same items. Each general strip in Figure 2(b) is
composed of multiple items of different sizes. -erefore, 2. Literature Review
there are four common types of staged patterns, including
two-staged homogenous patterns (2SHP), two-staged gen- To review the gradual advancement of the literature for
eral patterns (2SGP), three-staged homogenous patterns solving the 2DRCSP clearly, the related literature in the last
(3SHP), and three-staged general patterns (3SGP) in Fig- two decades is reviewed according to chronological order.
ure 3. -e arrows denote the position of a guillotine cut. -e -e critical analysis of the proposed methodology in each
numbers nearby the arrows denote the sequence of guillotine literature is also presented simultaneously.
cut in Figure 3. For example, as shown in Figure 3(a), the Hifi and Roucairol [3] presented an approximate algo-
first guillotine cut (GC_1) is horizontal to cut off the topside rithm and an exact algorithm for solving the 2DRCSP with
strip. -e second guillotine cut (GC_2) is vertical to separate restriction of two-staged guillotine cutting. -eir approxi-
three items. Generally, the 2SHP have the lowest material mate algorithm was constructed by using a dynamic pro-
utilization and simplest cutting complexity, and 3SGP have gramming procedure to solve multiple bounded knapsack
the opposite. -e 2SGP and 3SHP can compromise material problems. -eir exact algorithm was built by a tailored
utilization and cutting complexity. Meanwhile, all items branch-and-bound procedure, which determined lower and
designed on the plate can be divided exactly with three upper bounds by dynamic programming properties and
guillotine cuts according to the 2SGP and 3SHP, which may branched based on the solution space. As their proposed
be more suitable for a cutting machine during the cutting algorithms mainly apply dynamic programming, which is a
stock process. -erefore, we choose the 2SGP and 3SHP as typical greedy algorithm. -e solution by their proposed
candidate staged patterns for constructing a cutting plan in algorithms is easily trapped into local optimization.
this paper. -erefore, the proposed algorithms can be improved to get a
Staged patterns are necessary to satisfy guillotine cutting, better solution.
which is restricted by cutting machines during the Vanderbeck [4] focused on the 2DRCSP with restriction
manufacturing process of cutting flat glass in the glass in- of three-staged guillotine cutting. -e optimization objective
dustry, rocks in the granite and marble industries, steel focused on minimizing waste and other production goals,
blocks in the metallurgical industry, and so on. For instance, including aging stock, optional orders, and fixed setup cost.
the cutting process for steel plates is shown in Figure 4, -eir solution approach is a repeated use of the column
where a rectangular steel plate is cut into demanded items by generation method to get an approximate solution for the
three guillotine cuts. In the first guillotine cutting with single 2DRCSP. -is approach can get an optimal solution in
side crop shear, a rectangular steel plate is divided into acceptable computation time when the type number of order
segment 1 and segment 2. In the second guillotine cutting pieces is less. However, the solution space becomes huge
with double side crop shear, segment 1 is separated into when the type number of order pieces is large. It is hard to
strips 1, 2, and 3. Meanwhile, segment 2 is separated into achieve an optimal solution.
strips 4, 5, and 6. In the final guillotine cutting with multiple Alvarez-Valdes et al. [5] developed several heuristic
side crop shear, strips 1, 2, and 3 are all cut into two pieces of approaches for solving the 2DRCSP based on the framework
item 1. Strip 4 is cut into three pieces of item 1 and one waste. of the Gilmore–Gomory column generation scheme [6].
Strip 5 is cut into two pieces of item 3 and another waste. -ey use different heuristic algorithms, including the
Strip 6 is exactly cut into four pieces of item 2. Similar to the GRASP and tabu search, to generate each new cutting
paper cutting problem and the glass cutting problem, the pattern by solving the subproblem. It can increase solution
steel plate cutting problem has also been treated as the space at the expense of computation time. -eir proposed
application of the 2DRCSP with staged patterns. method actually belongs to column generation techniques,
Based on the 2SGP and 3SHP, a heuristic approach is which is suitable for 2DRCSP with less number of order
proposed to solve the 2DRSSSCSP-GR. Firstly, the 2SGP and piece type.
 5928, 2021, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2021/3732720 by Readcube (Labtiva Inc.), Wiley Online Library on [24/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Advances in Materials Science and Engineering 3

(a) (b) (c) (d)

Figure 1: Four variants of the 2DRSSSCSP: (a) 2DRSSSCSP-GO. (b) 2DRSSSCSP-GR. (c) 2DRSSSCSP-FO. (d) 2DRSSSCSP-FR.

3 3 5 5
2 2 2 2 2

2 2 4 4 4 4 4 4
1 1 1 1

(a) (b)

Figure 2: Two types of strips. (a) Homogenous strips. (b) General strips.

GC_2 GC_2 GC_2 GC_2 GC_2 GC_2

GC_3

GC_1 3 3 3 GC_1 3 3 5 5

2 2 2 2 2 2 2 2 4 4 4 4 4 4

1 1 1 1 1 1 1 1 1 1

(a) (b)
GC_3 GC_1 GC_3 GC_3 GC_3 GC_3 GC_1 GC_3 GC_3 GC_3
GC_4

1 1 1 1 1 1 1 1 1 6 6
GC_2 GC_2

3 3 3 3
1 1 1 1

2 2 2 2 2 2 2 2
1 1 1 1

(c) (d)

Figure 3: Four common types of staged patterns. (a) 2SHP. (b) 2SGP. (c) 3SHP. (d) 3SGP.

Suliman [7] proposed a sequential heuristic procedure can trap into locally optimal solution and hard to get globally
for the 2DRCSP to minimize the trim loss. -e sequential optimal solution.
heuristic procedure consisted of three steps. A width-cutting Belov and Scheithauer [8] solved the 2DRCSP with
pattern with the minimum width trim loss was obtained in restriction of two-staged guillotine cutting by applying a
the first step. A cutting pattern was achieved when the length branch-and-cut-and-price algorithm, which is a combina-
was determined with the minimum length trim loss. -e tion of the LP-based approach in [6] and branch-and-bound
frequency of the generated cutting pattern was confirmed with column generation in the Chvatal–Gomory cutting
based on the remaining number of each item. -e proposed planes. Based on their combination approach, all optimal
procedure is the simple sequential heuristic algorithm, solutions can be achieved for medium instances in ac-
which can get a solution in a short computation time. But it ceptable computation time. However, there were still
 5928, 2021, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2021/3732720 by Readcube (Labtiva Inc.), Wiley Online Library on [24/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
4 Advances in Materials Science and Engineering

single side
crop shear

steel plate segment 1 segment 2

double side
crop shear

strip 1 strip 4
Item 1 Item 1 Item 1 Item 1 Item 1
multiple side
crop shear
Item 3 Item 3 strip 5
Item 1 Item 1 strip 2

Item 2 Item 2 Item 2 Item 2 strip 6

Item 1 Item 1 strip 3

Figure 4: Cutting process for steel plates.

unsolved instances for larger instances because of the huge Silva et al. [12] introduced an integer programming
complexity of column generation with the large item type. model for the 2DRCSP with four restrictions: two-stage,
Cintra et al. [9] focused on solving the 2DRCSP with the three-stage, exact, and nonexact. -ey also addressed other
constraint of guillotine cutting. -eir approach framework constraints, such as the length of the cuts and the value of the
was also column generation. For generating each new remaining plate. -eir proposed model was solved by a
cutting pattern, they used two dynamic programming-based commercial solver (CPLEX). -eir conclusion claimed that
algorithms for the Rectangular Knapsack problem. -eir most instances with many item types could be solved in two
method also was not suitable for the instances with large hours. Several instances were dealt with over two hours.
item type. Obviously, it is a long computational time to get an optimal
Cui [10] proposed exact and heuristic algorithms to solve solution for the 2DRCSP occurring in the real enterprise
the 2DRCSP with three-staged guillotine cutting. -e exact application.
algorithm was actually a branch-and-bound approach, where Macedo et al. [13] presented an integer linear pro-
the lower and upper bounds were obtained by applying dy- gramming arc-flow model for the 2DRCSP with restriction
namic programming techniques to solve the knapsack prob- of two-staged guillotine cutting. -ey solved the model with
lems. Meanwhile, one heuristic algorithm was used to get the the commercial solver (CPLEX). -e computational results
initial lower bound for the main framework of the exact al- showed that the proposed arc-flow model also could not
gorithm. -e other heuristic algorithm underestimated the solve all the instances. Meanwhile, some instances needed a
upper bounds of the blocks generated for the main framework long computational time to get their solutions.
of the exact algorithm. -e proposed approach just only Cui and Huang [14] focused on the 2DRCSP with re-
considered homogenous strips, which was the basic element of striction of three-staged guillotine cutting. -e optimization
the cutting pattern. It can be capable of solving problems with objective was minimizing the sum of used plate cost and
large scale at the expense of reducing the solution space. cutting cost. -e problem was solved by the classical column
Hifi et al. [11] solved the 2DRCSP with restriction of generation framework, where the new cutting patterns were
two-staged guillotine cutting and fixed-orientation item. -e generated by a pattern-generation procedure. -ey
two-staged cutting patterns were generated by a strip-gen- researched the 2DRCSP with multiple objectives in a dif-
eration procedure. -eir proposed beam search was a ferent perspective.
truncated branch-and-bound algorithm, which selected the Cui [15] also addressed the 2DRCSP with three-staged
elite nodes based on an evaluation operator. -eir method patterns, which can be constructed in three steps. A set of
can solve the problems with large scale. But the two-staged rectangular items can be selected to generate strips in the
cutting pattern was a simple and special cutting pattern, first step. A set of generated strips can be chosen to build
whose material utilization was hard to be improved with segments in the second step. A set of generated segments can
iterative value correction. -e approach was also to get be selected to achieve cutting patterns in the third step. -ree
approximate solution by reducing solution space. large knapsack problems are constructed according to each
 5928, 2021, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2021/3732720 by Readcube (Labtiva Inc.), Wiley Online Library on [24/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Advances in Materials Science and Engineering 5

step. -e proposed dynamic programming algorithm was -e models that are in the approach were solved with the
used to solve each knapsack problem to get an optimal libraries of the CPLEX. -ey also cannot solve all instances
pattern. -e approach is simple and fast to get the solution. in acceptable computation time.
However, the greedy nature of dynamic programming may Velasco and Uchoa [23] considered the constrained two-
be influenced to get the optimal solution. dimensional guillotine cutting problem. -ey presented a
Cui and Zhao [16] presented a heuristic algorithm to weight adjusting algorithm based on integer programming,
solve the 2DRCSP with two-staged patterns. -e proposed which was solved using a dynamic programming state-space
algorithm was a special column generation method, which relaxation. -e solution can be obtained by using several
solved the residual problems repeatedly until all demands iterations of a GRASP-based primal heuristic. -e com-
were satisfied. Each pattern was generated by using a pro- putation time of their approach is not short to get an optimal
cedure for the constrained single large object placement solution.
problem. -ey combined the column generation and heu- Long et al. [24] solved the 2DRCSP with nonexact three-
ristic algorithm. It is a new idea to solve the 2DRCSP. stage patterns. Two additional practical constraints had
Cui et al. [17] focused on the 2DRCSP with the objective appeared in the cutting production line of steel mills. -ey
of plate minimization and pattern reduction. -ey proposed were the length between any two adjacent guillotine cuts and
a sequential grouping heuristic based on the framework of the number of guillotine cuts in the second stage by one.
the sequential heuristic procedure. It also applied a grouping Four heuristic algorithms based on column generation were
technique to select the items, which generated the next introduced to solve the practical 2DRCSP. -eir approach
pattern by using the dynamic programming algorithm. mainly included the column generation technique. Obvi-
Meanwhile, the item value can be adjusted with the se- ously, the computation time may be too long for some
quential value correction method. -eir approach was ac- instances with larger item type.
tually an improved sequential heuristic procedure. It can be Martin et al. [25] proposed a top-down cutting approach
suitable for solving the 2DRCSP with multiple optimization to solve the mixed-integer linear programming models for
objectives. the 2DRCSP with Guillotine cutting. -eir approach was
Cui et al. [18] also addressed the 2DRCSP with three- similar to the branch-and-bound algorithm. It used a binary
staged patterns. -ey proposed a heuristic algorithm tree to represent cutting patterns. -e root node was the
according to the construction process of the three-staged initial plate, and branches correspond to guillotine cuts.
pattern. -ey used the dynamic programming algorithm -ey also used a commercial solver (CPLEX) to solve the
three times to generate strips, segments, and patterns, re- established model.
spectively. -eir approach can get a solution in a short We can give the summary based on all reviewed liter-
computation time. But it was hard to get an optimal solution ature. -ere are mainly three optimization methods for
because of the greedy nature of the used algorithm. solving the 2DRCSP. -e first optimization method is based
Chen et al. [19] proposed a sequential value correction on the integer programming model, which is established
heuristic for solving the 2DRCSP with constrained three- according to the 2DRCSP with special constraints and solved
staged patterns. -eir main method was similar to the se- with branch-and-bound or existing commercial solver such
quential grouping heuristic proposed in Cui and Zhao [16]. as CPLEX. Based on [3, 10, 13, 20, 22, 25], the computational
-e difference was that Chen et al. [19] used three-staged complexity is too high with large item type. -e runtime of
homogenous patterns as the element for the cutting plan. solving several instances is even over two hours. -e second
Cui and Zhao [16] applied simple block patterns, otherwise. optimization method is based on a linear programming
Andrade et al. [20] solved the nonexact two-stage model, which is built with special constraints, such as the
2DRCSP with usable leftovers. -ey built two bilevel staged pattern. It can often be solved with the framework of
mathematical programming models, which can be refor- the column generation approach. Based on [4, 6, 8, 9, 14, 24],
mulated as one-level MIP models. -ey solved the models by the computational complexity becomes high with item type
applying the branch-and-cut method of CPLEX. increase. But the runtime of solving all instances is in two
Cui et al. [21] proposed a heuristic algorithm for the hours (acceptable computation time) at the expense of so-
2DRCSP with guillotine cutting. -eir algorithm included lution space, such as using staged patterns. -e third op-
three procedures. A pattern-generation procedure was used timization method is the heuristic procedure, whose main
to generate a set of triple-block patterns in the first pro- thought is to generate a cutting pattern using the remaining
cedure. A plan-generation procedure was applied to gen- items with some pattern-generation algorithm repeatedly
erate a solution in the second procedure. A plan- until all items demand is fulfilled. Based on [7, 15, 19, 21, 23],
improvement procedure is also used to generate an im- the computational complexity is moderate with large item
proved solution. -eir heuristic algorithm was also the same type. -e runtime of solving all instances is in a short time.
as the sequential value correction heuristic presented in According to the contrastive analysis, we select the third
Chen et al. [19]. method to solve the 2DRCSP. Meanwhile, our method is
Nascimento et al. [22] presented the 2DRCSP with different from other reviewed literature. We consider the
guillotine cutting. -ey considered the possibility of 2 and 3 combination of multiple staged patterns to solve the
cutting stages in a predefined sequence. A two-level iterative 2DRCSP. It is the first time to integrate two types of staged
approach was used to solve integer linear programming and patterns to solve the 2DRSSSCSP-GR. -erefore, this paper
constraint programming models for the special 2DRCSP. aims to build two types of staged patterns (2SGP and 3SHP)
 5928, 2021, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2021/3732720 by Readcube (Labtiva Inc.), Wiley Online Library on [24/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
6 Advances in Materials Science and Engineering

and present the ISHP approach combined with the 2SGP k � 1, 2, . . . , 2m. -e number of each item in type-k strip
and 3SHP in the remaining content. L ⊗ wk is determined by solving the following recursion:
f2SGP (k, x) � max􏼈f2SGP (k − 1, x), vi
3. Generation Procedures for 2SGP and 3SHP
+f2SGP k, x − lk 􏼁|lk ≤ x􏼉,
In the 2DRSSSCSP-GR, m types of rectangular items need to (1)
m
be cut off exactly from plates of the same size L ⊗ W f2SGP (k, x) � 􏽘 vi n2SGP (k, x, i), n2SGP (k, x, i) ≤ ri .
(length ⊗ width). -e type-i item can be characterized by the i�1
size li ⊗ wi ⊗ vi ⊗ di (length ⊗ width ⊗ value ⊗ demand). To
satisfy two restrictions (guillotine cuts and rotation) in the -e recursion is initialized with f2SGP (0, x) � 0 and
2DRSSSCSP-GR, the 2SGP and 3SHP are used to meet the n2SGP (0, x, i) � 0. -e value of type-k strip x ⊗ wk
restriction of guillotine cuts (G). 2m types of strips are (x � 1, 2, . . . , L) can be selected from the maximum value
considered to satisfy the restriction of the item allowed to be between the value of strip x ⊗ wk−1 and the other one, which
rotated by 90° (R). We firstly introduce generation proce- is the total value of one type-i item and strip (x − lk ) ⊗ wk .
dures for 2SGP and 3SHP. To simplify the presentation, only Meanwhile, the number of each item in type-k strip L ⊗ wk
X-patterns are presented in this section. -e Y-patterns are can be determined by n2SGP (k, x, i).
easy to be achieved by swapping the length and width of the All general strips can be determined by the decision
used plates. variable n2SGP (k, x, i), which is considered by
-e notations are used to denote the following related k � 1, 2, . . . , 2m, x � 1, 2, . . . , L,i � 1, 2, . . . , m. -erefore,
parameters: the complexity of recursion (1) is O(2m2 L).
-e general strips have been generated through a hor-
K: number of types of item allowed to be rotated by 90°, izontal combination of demanded items with the above
that is, K � 2m recursion. -e 2DGP for the plate L ⊗ W can be achieved
lk : length of type-k item, lk � li when k � 1, 2, . . . , m; through a vertical combination of generated strips with the
lk � wi when k � m + 1, m + 2, . . . , 2m, i � 1, 2, . . . , m following recursion:
wk : width of type-k item, wk � wi when k � 1, 2, . . . , m; F2SGP (y) � max􏼈F2SGP (y − 1), V2SGP (k)
wk � li when k � m + 1, m + 2, ..., 2m, i � 1, 2, . . . , m
+ F2SGP y − wk 􏼁|wk ≤y},
ri : remaining demand of type-i item m
f2SGP (k, x): value of type-k strip x ⊗ wk for 2SGP, V2SGP (k) � 􏽘􏼈vi min􏼂n2SGP (k, L, i), ri − N2SGP y − wk , i􏼁􏼃􏼉.
k � 1, 2, . . . , m i�1

n2SGP (k, x, i): number of type-i item in type-k strip (2)

x ⊗ wk for 2SGP, i � 1, 2, . . . , m
-is recursion is initialized with F2SGP (0) � 0 and
F2SGP (y): value of segmentL ⊗ y for 2SGP, N2SGP (0, i) � 0. -e value of segment L ⊗ y(y � 1, 2, . . . , W)
y � 1, 2, . . . , W can be chosen from the maximum value between the value of
V2SGP (k): total value of possible items in type-k strip segment L ⊗ (y − 1) and the other value, which is the total
L ⊗ wk for 2SGP value of strip L ⊗ wk and segment L ⊗ (y − wi ). Considering
N2SGP (y, i): number of type-i item in segment L ⊗ y for the remaining demand of each item, the value of type-k strip
2SGP, y � 1, 2, . . . , W L ⊗ wk can be computed by the sum of the price of each item
multiplying the possible existing number of each item, which
F3SHP (x, y): value of segmentx ⊗ y for 3SHP
is determined with the minimum between the number of
V3SHP (k, x): total value of possible existing items in strip L ⊗ wk and the remaining number of each item ex-
type-k strip x ⊗ wk for 3SHP cluding used for segment L ⊗ (y − wi ).
n3SHP (k, x, i): number of only including type-i item in -e 2DGP can be achieved with the maximum value
type-k strip x ⊗ wk for 3SHP, i � 1, 2, . . . , m F2SGP (y), which is considered by y � 1, 2, . . . , W. Mean-
N3SHP (x, y, i): number of type-i item in segment x ⊗ y while, the total value of type-k strip V2SGP (k) is determined
for 3SHP by n2SGP (k, L, i), which is considered by k � 1, 2, . . . , 2m,
i � 1, 2, . . . , m. -erefore, the complexity of recursion (2) is
T3SHP (x): value of 3SHP including segment x ⊗ W and
O(2m2 W). Subsequently, the complexity of the generation
segment (L − x) ⊗ W at the dividing position x
procedure for 2SGP is O[2m2 (L + W)].

3.1. Generation Procedure for 2SGP. -e generation proce- 3.2. Generation Procedure for 3SHP. -e generation proce-
dure for 2SGP always includes generation of general strips dure for 3SHP always includes the generation of homoge-
and generation of 2SGP. -e generation of general strips nous strips, homogenous segments, and 3SHP. -e
provides the candidate strips for 2SGP. Assume that all items generation of homogenous strips provides the candidate
are ordered by their width, that is, w1 ≤ w2 ≤ . . . ≤ w2m . 2m strips for homogenous segments. -e homogenous segments
types of general strips can be generated for 2SGP in the plate are the basic element for 3SHP. Similar to the generation
L ⊗ W, where the type-k strip has the size L ⊗ wk , procedure for 2SGP, the homogenous strips are constructed
 5928, 2021, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2021/3732720 by Readcube (Labtiva Inc.), Wiley Online Library on [24/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Advances in Materials Science and Engineering 7

along the length direction of the used plate L ⊗ W. Con- demanded items. -e variable is increased by one unit using
sidering item rotation, 2m type of homogenous strips x ⊗ wk general recursive techniques at each time, whereas only the
, including single item, are determined as follows: positions of normal sizes are considered to generate staged
patterns. Hence, considering normal sizes can reduce the
x
n3SHP (k, x, i) � min􏼨⌊ ⌋, ri 􏼩, k � 1, 2, . . . , m, computation time and cannot affect the solution quality. -e
li normal sizes of strips, segments, and patterns for T-shape
homogenous patterns are defined in Cui [26].
x Four sets of normal sizes are shown in formula (6). For
n3SHP (k, x, i) � min􏼨⌊ ⌋, ri 􏼩, k � m + 1, m + 2, . . . , 2m.
wi the generation procedure for 2SGP, the normal sizes from
(3) set X1 can be considered to generate strips x ⊗ wk when
recursion (1) is solved. -e normal sizes from set Y1 can be
-e homogenous segments can be obtained with a considered to generate patterns L ⊗ y when recursion (2) is
vertical combination of generated homogenous strips. -e solved. For the generation procedure for 3SHP, the normal
combination of homogenous strips for the segment x ⊗ W sizes from set X2 can be considered to generate strips x ⊗ wk
can be determined by solving the following recursion: when recursion (3) is solved. -e normal patterns from set
F3SHP (x, y) � max􏼈F3SHP (x, y − 1), V3SHP (k, x) Y2 can be considered to generate segments x ⊗ y when
recursion (4) is solved. -e normal patterns from set X2 can
+ F3SHP x, y − wk 􏼁|wk ≤ y􏼉, be considered to generate segments x ⊗ W when formula (5)
V3SHP (k, x) � vi min􏼂n3SHP (k, x, i), ri − N3SHP x, y − wk , i􏼁􏼃. is solved:
⎨
⎧ 2m ⎬
(4) L ⎫
X1 � ⎩ x|x � 􏽘 ek lk ≤ L, ek ∈ 􏼨0, 1, . . . , ⌊ ⌋􏼩⎭ ,
Recursion (4) is initialized with F3SHP (x, 0) � 0 and k�1
lk
N3SHP (x, 0, i) � 0. -e value of segment x ⊗ y can be
⎨
⎧ 2m ⎬
computed with the maximum value between the value of W ⎫
Y1 � ⎩ y|y � 􏽘 ek wk ≤ |W, ek ∈ 􏼨0, . . . , ⌊ ⌋􏼩⎭ ,
segment x ⊗ (y − 1) and the other value, which is the total w k
k�1
value of homogenous strip x ⊗ wk and segment x ⊗ (y − wi ). (6)
-e value of homogenous strip x ⊗ wk is the total value of L
including type-i items, which are selected from minimum X2 � 􏼨x|x � ek lk , ek ∈ 􏼨0, 1, . . . , ⌊ ⌋􏼩􏼩,
lk
between the number of homogenous strip x ⊗ wk , and the
remaining number of type-i item for the used segment
W
x ⊗ (y − wi ). Y2 � 􏼨y|y � ek wk , ek ∈ 􏼨0, 1, . . . , ⌊ ⌋􏼩􏼩.
-e generation of homogenous strips can be built with wk
formula (3). Based on all generated homogenous strips, the
homogenous segments can be generated by solving recur-
sion (4). -e decision variable is N3SHP (x, y, i) for 3.4. Correction Value for Each Item. -e recursive techniques
x � 1, 2, . . . , L, y � 1, 2, . . . , W, i � 1, 2, . . . , m. -e com- are used repeatedly during the generation procedure for
plexity of recursion (2) is O(mLW). 2SGP and 3SHP. -e solution may include good patterns in
-e 3SHP can be obtained with a horizontal combina- the early recursive stage and bad patterns in the later re-
tion of at least two segments. Although 3SHP including cursive stage because of the greedy nature. To avoid trapping
more segments can improve material utilization, the cutting into local optimization, the correction value is formula (7),
complexity is also increased. We present the simplest 3SHP, where the new value of each item can be changed according
including two segments, which can be extended to 3SHP, to those items in the current pattern. Let ni be the number of
including more segments. -e value of 3SHP is the total type-i item in the current pattern and u the material utili-
value of segment x ⊗ W and segment (L − x) ⊗ W, shown in zation of the current pattern:
formula (5). -e dividing position of plate L ⊗ W can be
li wi 􏼁
determined by selecting the maximum value of T3SHP (x): vi � g1 vi + g2 , i � 1, 2, . . . , m,
u
T3SHP (x) � max􏼈F3SHP (x, W) + F3SHP (L − x, W)􏼉, L ⊗ wk . αni
g1 � 1 − g2 , g2 � , (7)
(5) di + ri

-e formula is used to find the maximum value of Σm

i�1 ni li wi 􏼁
T3SHP (x), L ⊗ wk . -e complexity of formula (5) is O(L). u� .
LW
Considering that L ≪ mLW, subsequently, the complexity of
the generation procedure for 3SHP is O(mLW). -e new value of each item is determined with a
weighted combination of the existing value vi and the other
part (li wi )β /u, which is called overproportional material
3.3. Normal Sizes for 2SGP and 3SHP. -e staged patterns consumption norm in [27]. Two weighted parameters g1
with recursive techniques commonly consider the normal and g2 are computed by each computational formula. Pa-
sizes, which are the possible combination sizes of all rameter α is selected from interval [0.6, 0.9]. Parameter β is
 5928, 2021, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2021/3732720 by Readcube (Labtiva Inc.), Wiley Online Library on [24/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
8 Advances in Materials Science and Engineering

slightly bigger than 1 in [16]. -e value of the items non- -e procedure of the ISHP approach mainly includes the
existent in the current pattern will not be changed because following steps:
g2 � 0 and g1 � 1 when ni � 0.
Step 1: initialize the relevant parameters: G � 1,
For overproportional material consumption norm
Npl best � +∞, and vi � li wi , i � 1, 2, . . . , m.
(li wi )β /u, the numerator (li wi )β denotes that the value of
the item is overproportional to its area. -ese items with a Step 2: generate a cutting plan with a combination of
large area have a big new value. -e denominator u shows 2SGP and 3SHP.
that the value of these items constructing the current Step 2.1: let P �, F �, and ri � di , i � 1, 2, . . . , m
pattern with bad material utilization has a big change. Step 2.2: for the remaining demand ri of type-i item,
-ese items with hard combination possess a big new use generation procedure for 2SHP and 3SHP to
value. -erefore, the correction value for the item in the generate 2SHP and 3SHP based on normal sizes
current pattern is aimed to select those items with large Step 2.3: select one with a large value from 2SHP
area and hard combination. -ey are applied to generate and 3SHP in Step 2.2, and take it as the candidate
patterns in the early recursive stage. -is approach can pattern pj
reduce the greedy nature of recursive techniques and Step 2.4: compute the frequency of the current can-
improve solution quality. didate pattern pj by the formulafj � min􏼈􏼄ri /ni 􏼅, i �
1, 2, . . . , m}. Put pj and fj into P and F, respectively
4. An ISHP Approach Step 2.5: update the remaining demand by the formula
ri � ri − fj ni , i � 1, 2, . . . , m
-e proposed ISHP approach tries to achieve a cutting plan Step 2.6: if 􏽐m i�1 ri > 0, correct value of each item vi ,
with the minimum number of plates for solving the i � 1, 2, . . . , m. j � j + 1 and go to Step 2.2; otherwise,
2DRSSSCSP-GR. -e flowchart of the ISHP approach is record the number of used plate Npl cur for the current
shown in Figure 5. Based on data information of the cutting plan, and go to Step 3
2DRSSSCSP-GR, A 2SGP and a 3SHP are constructed with
Step 3: if Npl cur < Npl best , record the current cutting
the generation procedure for 2SGP (Section 3.1) and gen-
plan as the best cutting plan. Otherwise, ignore this
eration procedure for 3SHP (Section 3.2) based on normal
current cutting plan. -e number of iterations in-
sizes for 2SGP and 3SHP (Section 3.3), respectively. -e
creased by one; that is, G � G + 1.
pattern with a larger value is selected as the candidate
pattern. -e frequency of the candidate pattern is deter- Step 4: if G > Gmax , achieve the best cutting plan for the
mined according to the residual demand of each item. -e 2DRSSSCSP-GR. Otherwise, go to Step 2.
information of the candidate pattern and its frequency is -e whole steps run through G main iterations, each of
recorded. -e residual demand of each item is updated based which is the SHP in Step 2. Step 1 is used to initialize it-
on the information of the candidate pattern. A new cutting erations, the number of used plates, and pattern types for the
plan is generated when the demand for each item is zero. best cutting plan. -e 2SGP and 3SHP are generated for
Otherwise, the value of each item is corrected based on the constructing a cutting plan until all items demand is fulfilled
current candidate pattern with a correction value for each in Step 2. Step 2 includes generating a 2SGP and a 3SHP,
item (Section 3.4). -e newly generated cutting plan is used selecting the current candidate pattern, updating the items
to replace the previous cutting plan when the newly gen- demand, and correcting the value for each item. Step 3
erated cutting plan with the number of used plates is less compares the number of used plates of the current cutting
than those in the previous cutting plan. Otherwise, the newly plan and the previous cutting plan. -e iterations are run
generated cutting plan is discarded. Another newly gener- Gmax times and achieve the best cutting plan with the
ated cutting plan is constructed repeatedly when the number minimum number of used plates for solving the
of iterations is not reached. Otherwise, a cutting plan with a 2DRSSSCSP-GR.
minimum number of used plates is achieved for solving the AS a new cutting plan can be generated by each iteration,
current 2DRSSSCSP-GR. and the ISHP approach is used to generate Gmax cutting
-e following notations are used to conveniently de- plans. Meanwhile, the number of patterns in a cutting plan
scribe the procedure of the ISHP approach. can be estimated by H � (􏽐m i�1 li wi di )/(LW). -e com-
G: number of iterations, G � 1, 2, . . . , Gmax plexity of generation for 2SHP and 3SHP is O[2m2 (L +
W)] + O(mLW) � O[2m2 (L + W) + mLW] according to
Gmax : maximum number of iterations
Sections 3.1 and 3.2. -erefore, the complexity of the pro-
Npl best : number of used plates for the best cutting plan posed ISHP approach is O[(2m2 (L + W) + mLW)HGmax ].
Npl cur : number of used plates for the current cutting
plan
5. Computational Results
P: set of patterns
pj : type-j pattern -is section is presented in two parts. -e relevant pa-
F: set of frequency of patterns rameters are set in the first part. -e proposed ISHP ap-
proach is compared with one published algorithm with
fj : frequency of type-j pattern 2SGP in [16] and one published algorithm with 3SHP in [19]
 5928, 2021, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2021/3732720 by Readcube (Labtiva Inc.), Wiley Online Library on [24/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Advances in Materials Science and Engineering 9

Start

Initialize data information for the

2DRSSSCSP-GR

Construct a 2SGP Construct a 3SHP

Select the pattern with larger value as the

candidate pattern

Determine the frequency of the candidate

pattern

Correct the value of each item

Record information of the candidate pattern
based on current candidate pattern

Update demand of each item

No
whether demand of each item is zero

Yes

Generate a new cutting plan

whether the number of used plates are No

less than in previous cutting plan

Yes

Replace the previous cutting plan with

new generated cutting plan

No
whether number of iterations is reached

Yes

Achieve the cutting plan with minimum

number of used plates

Figure 5: Flowchart of the ISHP approach.

 5928, 2021, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2021/3732720 by Readcube (Labtiva Inc.), Wiley Online Library on [24/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
10 Advances in Materials Science and Engineering

in the second part. -ree types of heuristic algorithms are Table 1: Computational results of different parameter setting.
used to deal with two sets A and B, which are obtained from Constants Variables Ntot t
Silva et al. [12]. -e ISHP runs on an Acer laptop with
2.30Gz Inter Core i5 and 8 GB RAM. α � 0.6 252 0.7
α � 0.7 251 0.7
β � 1.02, Gmax � 100
α � 0.8 251 0.8
5.1. Parameter Setting. -e solution quality and computa- α � 0.9 252 0.7
tional time are determined by main parameters α, β, and β � 1.00 252 0.6
Gmax . -ree parameters are considered with α ∈ [0.6, 0.9], α � 0.7, Gmax � 100
β � 1.02 251 0.7
β ∈ [1.00, 1.08], and Gmax ∈ [1, 100]. -e ISHP is applied to β � 1.04 252 0.7
solve instances from set A. -e computational results are β � 1.06 252 0.8
shown in Table 1, where Ntot denotes the total used plates for Gmax � 1 264 0.1
all instances and t represents the average time in seconds for Gmax � 10 258 0.1
each instance. Gmax � 30 254 0.2
α � 0.7, β � 1.02
It can be seen from the computational results of Table 1 Gmax � 50 251 0.3
Gmax � 80 251 0.5
that (1) the two parameters α and β with the given intervals
Gmax � 100 251 0.7
have little effect on the solution quality and computation
time; (2) the solution quality tends to be stable when the
maximum iterations Gmax is larger than 50; and (3) the
Table 2: Computational results for set A.
computation time is very short for all instances from set
B. Hence, we set the value of three parameters: α � 0.7, Instances NLB NISHP NRCCG NSVCH
β � 1.02, and Gmax � 100. A1 23 Δ Δ Δ
A2 12 Δ 13 Δ
A3 8 Δ Δ Δ
5.2. Comparing the ISHP with Two Published Algorithms.
A4 5 Δ Δ Δ
-e published algorithm in [16] applied the repeated con- A5 4 Δ 5 Δ
strained CG (RCCG) to solve the 2DRSSSCSP-GR with CHL1 6 Δ 7 Δ
2SGP. -e other published algorithm in [19] used sequential CHL2 3 Δ Δ Δ
value correction heuristic (SVCH) to solve the 2DRSSSCSP- CHL5 3 Δ 4 4
GR with 3SHP. -e lower bound NLB of solution for each CHL6 5 Δ 6 6
instance is copied from [19]. Notations NISHP , NRCCG , and CHL7 6 Δ Δ Δ
NSVCH are the number of used plates using the ISHP, RCCG, CU1 12 Δ Δ Δ
and SVCH for solving two sets A and B, respectively. No- CU2 14 15 15 15
tation Δ represents that the solution is equal to the lower CW1 9 10 10 10
CW2 12 Δ 13 Δ
bound for each instance.
CW3 16 Δ Δ Δ
-ree algorithms are applied to solve set A of 30 small- Hchl2 6 Δ Δ Δ
scale instances. -e computational results for set A are Hchl3s 3 Δ Δ Δ
displayed in Table 2. RCCG and SVCH solve 19 and 26 Hchl4s 2 Δ 3 Δ
instances to their lower bounds, respectively. But the ISHP Hchl6s 5 Δ Δ Δ
solves 28 instances to their lower bounds. Only the instances Hchl7s 7 Δ Δ Δ
(CU2 and CW1) need one more plate. As one more plate is Hchl8s 2 Δ Δ Δ
used for a staged pattern, three guillotine cuts are added for Hchl9s 10 Δ Δ Δ
cutting complexity. -erefore, the comparison results in- HH 2 Δ Δ Δ
dicate that the proposed ISHP can achieve a better solution OF1 4 Δ Δ Δ
(higher material utilization and less cutting complexity) than OF2 4 Δ 5 Δ
STS2 12 Δ Δ Δ
the other two published algorithms for solving small-scale
STS4 5 Δ Δ Δ
instances. W 24 Δ Δ Δ
Meanwhile, three algorithms are also used to solve the set 2 2 Δ Δ Δ
B of 20 medium-scale instances. -e computational results 3 23 Δ 24 Δ
for set B are shown in Table 3. In the same comparison way,
RCCG and SVCH solve 11 and 16 instances to their lower
bounds, respectively. However, ISHP solves 18 instances to less cutting complexity) than the other two published al-
their lower bounds. Only the instances (APT39, APT47) gorithms for solving medium-scale instances.
need one more plate. In case of instance APT30, the ISHP -e data information of instance APT30 is obtained as
can reach the lower bound. But the other two published follows.
algorithms cannot reach the lower bound. -e data infor- -e used plates have the same size 927 ⊗ 152. -e number
mation of instance APT30 is generated in Table 4. -e of items types is 38, and the total number of all items is 192.
cutting plan figure for instance APT30 is shown in Figure 6. -e size and demand of each item are randomly selected from
-e comparison results also indicate that the proposed ISHP li ∈ [57, 371], wi ∈ [7, 47], and di ∈ [4, 6]. -e detailed data
can achieve a better solution (higher material utilization and information of all demand items is shown in Table 4.
 5928, 2021, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2021/3732720 by Readcube (Labtiva Inc.), Wiley Online Library on [24/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Advances in Materials Science and Engineering 11

Table 3: Computational results for set B.

Instances NLB NISHP NRCCG NSVCH
APT30 8 Δ 9 9
APT31 14 Δ Δ Δ
APT32 12 Δ 13 Δ
APT33 12 Δ 13 Δ
APT34 6 Δ Δ Δ
APT35 8 Δ Δ Δ
APT36 8 Δ Δ Δ
APT37 11 Δ 12 Δ
APT38 10 Δ 11 Δ
APT39 11 12 12 12
APT40 15 Δ 16 Δ
APT41 12 Δ Δ Δ
APT42 15 Δ Δ Δ
APT43 12 Δ 13 13
APT44 9 Δ Δ Δ
APT45 8 Δ Δ Δ
APT46 11 Δ Δ Δ
APT47 12 13 13 13
APT48 8 Δ Δ Δ
APT49 5 Δ Δ Δ

Table 4: Detailed data information of all demand items.

ID li wi di
1 273 44 6
2 302 7 5
3 77 23 5
4 304 21 5
5 216 37 5
6 67 38 6
7 124 13 5
8 189 26 4
9 318 24 6
10 320 36 4
11 86 35 6
12 322 36 5
13 318 17 5
14 189 33 4
15 269 32 4
16 81 12 5
17 169 10 6
18 305 26 6
19 266 45 6
20 319 18 5
21 223 29 5
22 48 15 6
23 284 36 5
24 311 16 4
25 230 26 4
26 255 34 4
27 251 42 5
28 160 45 4
29 223 28 5
30 90 11 5
31 239 12 5
32 47 16 6
33 124 40 6
34 51 16 6
35 67 39 4
36 276 15 4
37 235 44 6
38 136 20 5
 5928, 2021, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2021/3732720 by Readcube (Labtiva Inc.), Wiley Online Library on [24/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
12 Advances in Materials Science and Engineering

9 25 33 26
9 25
9 25 33 26
12 23 20
20
12 23 20
20
38 38 38 6
10
37 37
10 6
37 37
10 6 1 1
37 37 1 1
10 6

18 18 18
5 5 12 11 11

1 1 27 33

28 28 19 1 35

18 18 18
5 5 12 11 11

1 1 27 33

28 28 19 1 35

3 24 5 3 22
27
3 9
222
3 9 21
27 3 22
3 13 21
13 4 22
16 13 26 21
16 3 22
16
16 12 26 29 4 22
16

7 22 31 17
7 31 21
4
7 15
7 4 21
7 4 15
4 29
33
4 15 29
33 24 15
24 29

31 31 31 30 30
34 34 32 32 32 32 32 24 36
25 8 8 9
36 36 36 30
19 19 27 35 35

6 23 23 23

30 30
8 8 20
2 2 2
38 38 13 13
14 14 14 29 3 34
19 19 6 14 34
17 17 17 17 17
Figure 6: A cutting plan figure for instance APT30 solved by the ISHP approach.
 5928, 2021, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2021/3732720 by Readcube (Labtiva Inc.), Wiley Online Library on [24/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Advances in Materials Science and Engineering 13

Table 5: Computational results of three approaches. selected as the best cutting plan for the current 2DRSSSCSP-
ID UISHP (%) URCCG (%) USVCH (%) NISHP NRCCG NSVCH
GR. -e computational results also indicate that the ISHP
approach is more effective in material utilization and cutting
Set A 92.79 89.58 92.07 53 55 53 complexity than the published algorithms with staged
Set B 93.33 89.66 92.28 143 150 145 patterns.
-e common staged patterns are welcomed in the
-e cutting plan figure for instance APT30 solved by the cutting process of the manufacturing industry, such as the
ISHP approach is shown in Figure 6, where the number production of steel products, paper, wood, glass, and other
indexes from 1 to 38 in the filled blue box represents “ID” of industries. 2SGP and 3SHP can be easily obtained with
each item in Table 4. -e filled slash box represents useable recursive techniques. -e ISHP with the combination of
leftovers. 2SGP and 3SHP can get the most instances to their lower
bounds, which verifies that combining 2SGP and 3SHP can
balance material utilization and cutting complexity while
5.3. Influence of Combining 2SGP and 3SHP. -is section solving the 2DRSSSCSP-GR.
analyses the influence of combining 2SGP and 3SHP for the -e ISHP in this paper considers the 2DRSSSCSP-GR
optimization objectives, including material utilization and with two contradictory objectives of material utilization and
cutting complexity. For generating a staged pattern, 2SGP cutting complexity. Future research may consider other
and 3SHP are used in our proposed ISHP approach, si- factors, such as setup cost or useable leftovers.
multaneously. -e only 2SGP is applied by the RCCG
presented in [16]. Only 3SHP is applied by the SVCH in [19]. Abbreviations
-ree methods are all used to solve all instances in set A and
set B. -e computational results of the three approaches are 2SGP: Two-staged general patterns
shown in Table 5. 3SHP: -ree-staged homogenous patterns
In Table 5, let U and N denote the average material 2DRCSP: Two-dimensional rectangular cutting stock
utilization and the average number of cuts, respectively. It is problem
noted that the average number of cuts is upper rounding. For 2DRSSSCSP- Two-dimensional rectangular single stock
example, Notations UISHP (%) and NISHP represent the av- GR: size cutting stock problem with guillotine
erage material utilization and the average number of cuts for cuts and item rotated by 90°
all instances in set A solved by the ISHP approach. Other ISHP: Iterative sequential heuristic procedure
parameters are the corresponding meaning. L ⊗ W: Plates of length and width
From a material utilization perspective, our proposed li ⊗ wi ⊗ vi ⊗ di : Type-i item of length, width, value, and
ISHP approach can get higher average material utilization demand
than the RCCG and SVCH approaches for solving instances K: Number of types of item allowed to be
in both set A and set B. -is verifies that our proposed ISHP rotated by 90°; that is K � 2m
approach is more effective in improving material utilization. lk : Length of type-k item, lk � li when
From a cutting complexity perspective, our proposed k � 1, 2, . . . , m; lk � wi when
ISHP approach can obtain the same average number of cuts k � m + 1, m + 2, . . . , 2m, i � 1, 2, . . . , m
as the SVCH approach, less than the RCCG approach for wk : Width of type-k item, wk � wi when
solving instances in set A. Or proposed ISHP approach can k � 1, 2, . . . , m; wk � li when
achieve less than the RCCG and SVCH approaches for k � m + 1, m + 2, . . . , 2m, i � 1, 2, . . . , m
solving instances in set B. -is indicates that our proposed ri : Remaining demand of type-i item
ISHP approach also can reduce cutting complexity without f2SGP (k, x): Value of type-k strip x ⊗ wk for 2SGP,
affecting material utilization. k � 1, 2, . . . , 2m
-erefore, our proposed ISHP approach considering the n2SGP (k, x, i): Number of type-i item in type-k strip x ⊗ wk
combination of 2SGP and 3SHP can get better solution for 2SGP i � 1, 2, . . . , m
quality than two other published approaches considering F2SGP (y): Value of segmentL ⊗ y for 2SGP
only 2SGP or 3SHP. y � 1, 2, . . . , W
V2SGP (k): Total value of possible items in type-k strip
6. Conclusions L ⊗ wk for 2SGP
N2SGP (y, i): number of type-i item in segment L ⊗ y for
-e ISHP approach in this paper is proposed to solve the 2SGP y � 1, 2, . . . , W
2DRSSSCSP-GR with the combination of 2SGP and 3SHP. F3SHP (x, y): Value of segmentx ⊗ y for 3SHP
-e 2SGP and 3SHP are as candidate staged patterns. -e V3SHP (k, x): Total value of possible existing items in
SHP algorithm is used to select a candidate staged pattern type-k strip x ⊗ wk for 3SHP
with a larger value. A new cutting plan is generated until all n3SHP (k, x, i): Number of only including type-i item in
items demand is fulfilled accurately. -e ISHP approach tries type-k strip x ⊗ wk for 3SHP, i � 1, 2, . . . , m
to generate many cutting plans by the iterative algorithm. N3SHP (x, y, i): Number of type-i item in segment x ⊗ y for
-e cutting plan with the minimum number of used plates is 3SHP
 5928, 2021, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2021/3732720 by Readcube (Labtiva Inc.), Wiley Online Library on [24/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
14 Advances in Materials Science and Engineering

dimensional two-stage cutting,” European Journal of Opera-

T3SHP (x): Value of 3SHP including segment x ⊗ W
tional Research, vol. 171, no. 1, pp. 85–106, 2006.
and segment (L − x) ⊗ W at the dividing [9] G. F. Cintra, F. K. Miyazawa, Y. Wakabayashi, and
position x E. C. Xavier, “Algorithms for two-dimensional cutting stock
α and β: Two adjustable parameters for correction and strip packing problems using dynamic programming and
value for each item column generation,” European Journal of Operational Re-
G: Number of iterationsG � 1, 2, . . . , Gmax search, vol. 191, no. 1, pp. 61–85, 2008.
Gmax : Maximum number of iterations [10] Y. Cui, “Heuristic and exact algorithms for generating ho-
Npl best : Number of used plates for the best cutting mogenous constrained three-staged cutting patterns,” Com-
plan puters & Operations Research, vol. 35, no. 1, pp. 212–225,
2008.
Npl cur : Number of used plates for the current
cutting plan [11] M. Hifi, R. M’Hallah, and T. Saadi, “Algorithms for the
constrained two-staged two-dimensional cutting problem,”
P: Set of patterns
INFORMS Journal on Computing, vol. 20, no. 2, pp. 212–221,
pj : Type-j pattern 2008.
F: Set of frequency of patterns [12] E. Silva, F. Alvelos, and J. M. Valério de Carvalho, “An integer
fj : Frequency of type-j pattern. programming model for two- and three-stage two-dimen-
sional cutting stock problems,” European Journal of Opera-
Data Availability tional Research, vol. 205, no. 3, pp. 699–708, 2010.
[13] R. Macedo, C. Alves, and J. M. Valério de Carvalho, “Arc-flow
-e data used to support the findings of this study are model for the two-dimensional guillotine cutting stock
available from the corresponding author upon request. problem,” Computers & Operations Research, vol. 37, no. 6,
pp. 991–1001, 2010.
[14] Y. Cui and B. Huang, “Reducing the number of cuts in
Conflicts of Interest generating three-staged cutting patterns,” European Journal of
Operational Research, vol. 218, no. 2, pp. 358–365, 2012.
-e authors declare that there are no conflicts of interest
[15] Y. Cui, “A new dynamic programming procedure for three-
regarding the publication of this paper. staged cutting patterns,” Journal of Global Optimization,
vol. 55, no. 2, pp. 349–357, 2013.
Acknowledgments [16] Y. Cui and Z. Zhao, “Heuristic for the rectangular two-di-
mensional single stock size cutting stock problem with two-
-is work was supported by the National Key Research and staged patterns,” European Journal of Operational Research,
Development Program of China (grant nos. vol. 231, no. 2, pp. 288–298, 2013.
2018YFB0105300 and 2017YFD0700600). [17] Y. Cui, L. Yang, Z. Zhao, T. Tang, and M. Yin, “Sequential
grouping heuristic for the two-dimensional cutting stock
problem with pattern reduction,” International Journal of
References Production Economics, vol. 144, no. 2, pp. 432–439, 2013.
[1] G. Wäscher, H. Haussner, and H. Schumann, “An improved [18] Y.-P. Cui, Y. Cui, T. Tang, and W. Hu, “Heuristic for con-
typology of cutting and packing problems,” European Journal strained two-dimensional three-staged patterns,” Journal of
of Operational Research, vol. 183, no. 3, pp. 1109–1130, 2007. the Operational Research Society, vol. 66, no. 4, pp. 647–656,
[2] A. Lodi, S. Martello, and D. Vigo, “Heuristic and meta- 2015.
heuristic approaches for a class of two-dimensional bin [19] Q. Chen, Y. Cui, and Y. Chen, “Sequential value correction
packing problems,” INFORMS Journal on Computing, vol. 11, heuristic for the two-dimensional cutting stock problem with
no. 4, pp. 345–357, 1999. three-staged homogenous patterns,” Optimization Methods
[3] M. Hifi and C. Roucairol, “Approximate and exact algorithms and Software, vol. 31, no. 1, pp. 68–87, 2016.
for constrained (un) weighted two-dimensional two-staged [20] R. Andrade, E. G. Birgin, and R. Morabito, “Two-stage two-
cutting stock problems,” Journal of Combinatorial Optimi- dimensional guillotine cutting stock problems with usable
zation, vol. 5, no. 4, pp. 465–494, 2001. leftover,” International Transactions in Operational Research,
[4] F. Vanderbeck, “A nested decomposition approach to a three- vol. 23, no. 1/2, pp. 121–145, 2016.
stage, two-dimensional cutting-stock problem,” Management [21] Y. P. Cui, Y. Yao, and D. Zhang, “Applying triple-block
Science, vol. 47, no. 6, pp. 864–879, 2001. patterns in solving the two-dimensional bin packing prob-
[5] R. Alvarez-Valdes, A. Parajon, and J. M. Tamarit, “A com- lem,” Journal of the Operational Research Society, vol. 69, 2017.
putational study of LP-based heuristic algorithms for two- [22] O. Nascimento, T. Queiroz, and L. Junqueira, “A MIP-CP
dimensional guillotine cutting stock problems,” Spectrum, based approach for two- and three-dimensional cutting
vol. 24, no. 2, pp. 179–192, 2002. problems with staged guillotine cuts,” Annals of Operations
[6] P. C. Gilmore and R. E. Gomory, “Multistage cutting stock Research, Article ID 214313095, 2019.
problems of two and more dimensions,” Operations Research, [23] A. S. Velasco and E. Uchoa, “Improved state space relaxation
vol. 13, no. 13, pp. 94–120, 1965. for constrained two-dimensional guillotine cutting prob-
[7] S. M. A. Suliman, “A sequential heuristic procedure for the lems,” European Journal of Operational Research, vol. 272,
two-dimensional cutting-stock problem,” International no. 1, pp. 106–120, 2019.
Journal of Production Economics, vol. 99, no. 1, pp. 177–185, [24] J. Long, Z. Zheng, X. Gao, P. M. Pardalos, and W. Hu, “An
2006. effective heuristic based on column generation for the two-
[8] G. Belov and G. Scheithauer, “A branch-and-cut-and-price dimensional three-stage steel plate cutting problem,” Annals
algorithm for one-dimensional stock cutting and two- of Operations Research, vol. 289, 2020.
 5928, 2021, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1155/2021/3732720 by Readcube (Labtiva Inc.), Wiley Online Library on [24/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Advances in Materials Science and Engineering 15

[25] M. Martin, R. Morabito, and P. Munari, “A top-down cutting

approach for modeling the constrained two- and three-di-
mensional guillotine cutting problems,” Journal of the Op-
erational Research Society, no. 11, pp. 1–15, 2020.
[26] Y. Cui, “An exact algorithm for generating homogenous
T-shape cutting patterns,” Computers & Operations Research,
vol. 34, no. 4, pp. 1107–1120, 2007.
[27] G. Belov and G. Scheithauer, “Setup and open-stacks mini-
mization in one-dimensional stock cutting,” INFORMS
Journal on Computing, vol. 19, no. 1, pp. 27–35, 2007.