One dimensional cutting stock problem 1-d-csp_ a study on data dependent tri

International Journal of Computer Engineering and Technology (IJCET), ISSN 0976-
6367(Print), ISSN 0976 – 6375(Online) Volume 4, Issue 4, July-August (2013), © IAEME
524
ONE-DIMENSIONAL CUTTING STOCK PROBLEM (1D-CSP): A STUDY ON
DATA DEPENDENT TRIM LOSS
P. L. Powar1
, Vinit Jain2
, Manish Saraf3
, Ravi Vishwakarma4
1
Deptt. of Math. & Comp. Sc., R. D. University, Jabalpur 482001, India
2
KEC Int. Company, Panagar, Jabalpur, 482001, India
3
HCET, Dumna Airport Road, Jabalpur, 482001, India
4
Deptt. of Math. & Comp. Sc., R. D. University, Jabalpur 482001, India
ABSTRACT
A generalized approach is introduced for getting minimal trim in one- dimensional cutting
stock problem (1D-CSP) which occurs extensively while manufacturing of different engineering
objects. Our study is especially focused on transmission tower manufacturing industry. The concept
of computing total trim by using the idea of pre-defined sustainable trim has been explored in
previous work of authors. Given m stock lengths ܷଵ, ܷଶ, … , ܷ௠, the cutting plan consists of cutting of
at most two order lengths at a time out of the demanded set of n order lengths ݈ଵ, ݈ଶ, … , ݈௡.
Considering the given data, the total trim has been computed corresponding to two different
sustainable trims of order one (viz.‫ݐ‬௦
ଵ
) and order two (viz.‫ݐ‬௦
ଶ
). Introducing various values of
sustainable trims as knots between ‫ݐ‬௦
ଵ
and ‫ݐ‬௦
ଶ
, the total trim has been computed corresponding to
each knot, the linear approximation has been constructed which predict the total trim loss at any
arbitrary point ‫ݐ‬ lying between ‫ݐ‬௦
ଵ
and ‫ݐ‬௦
ଶ
.
KEYWORDS: First order sustainable trim, Second order sustainable trim, 1D-CSP, Non-negative
integral valued (NIV) linear combination, Hat function, Linear approximation.
AMS (2000) SUBJECT CLASSIFICATION: 90C90; 90C27; 90C10.
1. INTRODUCTION
Various sizes of items are ordered in industries dealing with processing of materials such as
steel, film, textiles etc. The ordered items are cut from bars, plates or sheets of raw materials that
have a fixed size (cf. [10], [14], [15], [16]). During the process of cutting of these items, the problem
for trying to minimize the waste of raw material is referred to as a cutting stock problem (CSP). The
CSP plays an important role in the economic status of the processing industries and it has been
INTERNATIONAL JOURNAL OF COMPUTER ENGINEERING &
TECHNOLOGY (IJCET)
ISSN 0976 – 6367(Print)
ISSN 0976 – 6375(Online)
Volume 4, Issue 4, July-August (2013), pp. 524-538
© IAEME: www.iaeme.com/ijcet.asp
Journal Impact Factor (2013): 6.1302 (Calculated by GISI)
www.jifactor.com
IJCET
© I A E M E

525
entered into several fields of research like mathematical modeling, computer programming, ocean
engineering, mechanical engineering etc. (cf. [6], [11], [12], [16]). In the clothing industry, the first
phase of the stock cutting process is one-dimensional cutting of a smaller number of long pieces in
the form of fabric rolls into a large number of short pieces named pattern shapes (cf. [10]).
The analytic approaches of optimization have been proposed by several researchers (see [1],
[2], [4], [5], [7], [8], [9], [12], [13], [17], [18]), but it has been noticed that from implementation
point of view, these methods have been discarded by almost all industries due to their complexity.
The analytic approach consists of following consecutive steps:
• Mathematical modeling
• Algorithm
• Cutting plan
• Computation of trim loss
These steps have been proposed and verified theoretically and then the relevant industries are
supposed to adopt the method practically. But, it has been observed that there is a big-gap between
analytic solution and its practical implementation. Our work in this paper is mainly focused on the
transmission tower manufacturing industry and in accordance with the practical approach,
mathematical model has been designed. Powar et al in [12] (cf. [11] also) have proposed the more
functional cutting plan which can be conveniently acceptable by any industry dealing with 1D-CSP.
The cutting plan described in [12] (see also [11]) consists of cutting of at most two order
lengths at a time out of the required n order lengths ݈ଵ, ݈ଶ, … ݈௡, in from a given set of m stock lengths
ܷଵ, ܷଶ, … ܷ௠. At each stage, the cutting of at most two order lengths at a time will be continued till
the required number of pieces are cut totally. The concept of sustainable trim of order one ‫ݐ‬௦
ଵ
and of
order two ‫ݐ‬௦
ଶ
has been introduced in [12] and [11] respectively. This idea of pre-defined sustainable
trims plays a crucial role in controlling the total trim.
Spline functions (piece-wise polynomial functions) are well-known best approximating
functions which are extensively use in all branches of science and technology. For our analysis, we
have considered a clan of data and computed the weighted average of order lengths. Correspond to
each weighted average in the domain, the trim losses with respect to ‫ݐ‬௦
ଵ
and ‫ݐ‬௦
ଶ
have been computed.
Using the basis functions (Hat functions or Chapeau functions) for linear approximation, linear
splines (cf. [3]) are generated given any arbitrary set of data and the value of approximate trim loss
may be predicted immediately in accordance with our proposed cutting plan.
2. PRE-REQUISITES
All stock lengths and order lengths, we consider as integers throughout our analysis.
According to the requirement, the lengths can be converted into integers by multiplying them with
10௡
ሺ݊ ൒ 1, integerሻ. We use the following notations:
‫ܤ‬ሺ݊ሻ െ Block of integers 0,1, … , ݊ (index set), ݆ ‫א‬ ‫ܤ‬ሺ݊ሻ means ݆ can be any number from the set
ሼ0, 1, 2, … , ݊ሽ.
݈௜ െ Order lengths ݅ ൌ 0, 1, 2, … , ݊ arranged in ascending order with respect to length and ݈଴ ൌ 0 by
convention.
݀௜ െ Required number of pieces of order length ݈௜, ݀଴ ൌ 0.
ܷ௝ െ Stock length ሺ݆ ൌ 1, 2, … , ݉ሻ arranged in ascending order with respect to length.
It has been noticed that in particular, in the transmission tower designing industry that most
of the required number of order lengths i.e. ݀௜′s are integral multiple of each other. In view of this
observation, we classify the order lengths in the following two categories in accordance with their
required number of pieces:
Category I: (C-I) We collect all those order lengths whose required number of pieces are integral
multiple of each others.

526
Category II: (C-II) It is the collection of all those order lengths whose required number of pieces
are prime numbers (their common multiple is 1).
3. SUSTAINABLE TRIMS
3.1 Order One
In order to cut the linear combination ܽ௜௝ (say) of the two order lengths ݈௜ and ݈௝ from the
given stock lengths ܷଵ, ܷଶ, … , ܷ௠, we have to decide upto what extent, we allow the raw material to
convert into the scrape. Throughout our cutting process (excluding the last step where it is possible
only that few piece of some order length are left to cut), we follow the restriction that 0 ൑ ܷ௞ െ
ܽ௜௝ ൑ ‫ݐ‬௦
ଵ
, ݇ ൌ 1,2, … , ݉ and ‫ݐ‬௦
ଵ
is the sustainable trim of order one and defined as follows:
‫ܮ‬ ൌ
݈଴݀଴ ൅ ݈ଵ݀ଵ ൅ ‫ڮ‬ ൅ ݈௡݀௡
݀଴ ൅ ݀ଵ ൅ ‫ڮ‬ ൅ ݀௡
ൌ
∑ ݈௜݀௜
௡
௜ୀ଴
∑ ݀௝
௡
௝ୀ଴
We next define ‫ܮ‬௞ ൌ |ܷ௞ െ ݅‫|ܮ‬ (݇ ൌ 1, 2, … , ݉ and ݅ is an appropriate positive integer ൒
1, for which ‫ܮ‬௞ is minimum)
where ܷଵ, ܷଶ, … , ܷ௠ are the stock lengths. We finally define
‫ݐ‬௦
ଵ
ൌ
∑ ௅ೖ
೘
ೖసభ
௠
(3.1)
which is the desired sustainable trim of order one.
3.1 Remark
Analytically, it has been noticed that the average value covers the acceptable, over all
original values. Hence, we have taken the weighted mean of total required lengths.
3.2 Order Two
Following the same restriction as for ‫ݐ‬௦
ଵ
and using the notations from section 2, we define
ܵ௡ ൌ
l଴d଴ ൅ ‫ڮ‬ ൅ l୬d୬
d଴ ൅ dଵ ൅ ‫ڮ‬ ൅ d୬
ൌ
∑ ݈௜݀௜
௡
௜ୀ଴
∑ ݀௝
௡
௝ୀ଴
By convention, ܵ଴ ൌ 0, ܵଵ ൌ ݈ଵ, ܵଶ ൌ
௟బௗబା௟భௗభା௟మௗమ
ௗబାௗభାௗమ
, … , ܵ௡ ൌ
∑ ௟೔ௗ೔
೙
೔సబ
∑ ௗೕ
೙
ೕసబ
We next define the second order weighted means
ܵ௡
ଵ
ൌ
ܵ଴ ൅ ܵଵ ൅ ‫ڮ‬ ൅ ܵ௡
݊ ൅ 1
.
Consider ‫ܮ‬௞ ൌ |ܷ௞ െ ݅ ܵ௡
ଵ|, ݇ ൌ 1,2, … , ݉.
݅ is an appropriate positive integers ൒ 1, for which ‫ܮ‬௞ is minimum. ܷଵ, ܷଶ , … , ܷ௠ are stock lengths.

527
3.2 Remark
ܵ௡
ଵ
is the average order length which is assumed to be cut from the given stock length
ܷ௝ሺ݆ ൌ 1, … , ݉ሻ. The integer ݅ denotes the number of pieces of average stock length to be cut from
the stock length ܷ௞. We finally define
‫ݐ‬௦
ଶ
ൌ
∑ ௅ೖ
೙
ೖసభ
௠
(3.2)
which is the sustainable trim of order two viz. ‫ݐ‬௦
ଶ
.
4. MATHEMATICAL MODELING
We first consider C-I and define the following ratios:
ௗ೔
ௗೕ
ൌ
ఈ೔ൈ௥೔ೕ
ఉೕൈ௥೔ೕ
(4.1)
(where ‫ݎ‬௜௝ is a positive integer , ݆ ‫א‬ ‫ܤ‬ሺ݊ሻ, ݅ ൐ ݆ )
Note: It is not necessary to consider always the largest common factor between ݀௜ and ݀௝. Any other
factor ‫ݎ‬௜ (if exists) may be selected according to the length of stock to minimize the trim.
In view of (4.1), define the following set:
‫ܣ‬ ൌ ሼܽ௜௝ ൌ ߙ௜݈௜ ൅ ߚ௝݈௝ ‫׷‬ ܽ௜௝ ൑ ܷ௠, ߙ௜, ߚ௝ ൒ 0, integer ൫݅ ൏ ݆, ݅, ݆ ‫א‬ ‫ܤ‬ሺ݊ሻ൯ሽ (4.2)
We are now in a position to define the sets ‫ܣ‬௞ ‫ك‬ ‫ܣ‬ ሺ݇ ൌ 1, 2, … , ݉ሻ as follows:
‫ܣ‬௞ ൌ ሼܽ௜௝: 0 ൑ ܷ௞ െ ܽ௜௝ ൑ ‫ݐ‬௦
ఒ
, ݇ ‫א‬ ‫ܤ‬ሺ݉ሻ, ݅, ݆ ‫א‬ ‫ܤ‬ሺ݊ሻ, ܽ௜௝ ‫א‬ ‫,ܣ‬ ߣ ൌ 1,2ሽ (4.3)
where ‫ݐ‬௦
ఒ
is defined by (3.1) and (3.2) respectively for ߣ ൌ 1 and λ ൌ 2.
At this stage, we may come across with the following situations:
• ‫ܣ‬௞ ൌ ߶, ‫׊‬ ݇ ‫א‬ ‫ܤ‬ሺ݉ሻ, in this case, all the order lengths have to shift in C-II.
• In view of the definition of ‫ݐ‬௦
ఒ
, the sets ‫ܣ‬௞ ሺ݇ ‫א‬ ‫ܤ‬ሺ݉ሻሻ may or may not cover all order lengths
belonging to Category-I.
In view of above observations and the definition of the sets ‫ܣ‬௞, we redefine our categories I
and II as follows:
Category-I (C-I) Let ݈ఈଵ
, ݈ఈమ
, … , ݈ఈ௣
order lengths have been covered by the sets ‫ܣ‬௞ ሺ݇ ൌ
1, 2, . . , ݉ሻ. For convenience, we denote these order lengths by ݈ଵ, ݈ଶ, … , ݈௣ arranged in ascending
order with respect to the length.
4.1 Remark
There may exist some order lengths ݈௜ and ݈௝ (say) such that ݀௜ and ݀௝ ofcourse are multiple
of each others but the length of combination ܽ௜௝ exceeds the largest stock length ܷ௠ or ሺܷ௝ െ ܽ௜௝ሻ
exceeds the sustainable trim loss. We shift all such order lengths to Category-II and finally, we
assume that the order lengths ݈ଵ, ݈ଶ, … , ݈௣ have been covered by Category-I.
Category-II (C-II) The remaining all order lengths ܰ ൌ ݊ െ ‫݌‬ denoted by ݈ଵ, ݈ଶ, … , ݈ே arranged in
ascending order with respect to the length.

528
4.2 Remark
(i) The real numbers ‫ݐ‬௦
ఒ
ሺߣ ൌ 1, 2ሻ defined by ሺ3.1ሻ and ሺ3.2ሻ play a crucial role in the
computation of total trim loss. It is natural to expect that the trim loss can be minimized by
considering the minimum value lying between 0 and ‫ݐ‬௦
ఒ
, but it has been experienced practically
in the industries that by increasing the value of ‫ݐ‬௦
ఒ
, the impact on the total trim loss results in a
significantly acceptable range in some particular cases. But we are strict to ‫ݐ‬௦
ఒ
only.
(ii) In order to implement the algorithm smoothly, the data of more than one tower (preferably of
same pattern) may be clubbed.
Now consider Category-II and order lengths ݈ଵ, ݈ଶ, … , ݈ே with the required number of pieces
݀ଵ, ݀ଶ, … , ݀ே respectively. For ݅ ് ݆, ݅, ݆ ‫א‬ ‫ܤ‬ሺܰሻ, define:
݀௜ ൌ ݊ଵߙ௜ଵ ൅ ݀௜ଵ (4.4)
݀௝ ൌ ݊ଵߚ௝ଵ ൅ ݀௝ଵ (4.5)
0 ൑ ܷ௞ െ ൫݈௜ߙ௜ଵ ൅ ݈௝ߚ௝ଵ൯ ൌ ሺ‫ݓ‬௞ sayሻ ൑ ‫ݐ‬௦
ఒ
ሺߣ ൌ 1, 2ሻ
for at least one value of k (k=1,2,…,m). The number ݊ଵ has been chosen in such a way that ‫ݓ‬௞
attains a minimum value lying between 0 and ‫ݐ‬௦
ఒ
.
Similarly, choose a number ݊ଶ satisfying the following condition:
݀௜ଵ ൌ ݊ଶߙ௜ଶ ൅ ݀௜ଶ (4.6)
݀௝ଵ ൌ ݊ଶߚ௝ଶ ൅ ݀௝ଶ (4.7)
Proceeding this way, we finally define
݀௜,௦ିଵ ൌ ݊௦ߙ௜௦ ൅ ݀௜௦ (4.8)
݀௝,௦ିଵ ൌ ݊௦ߚ௝௦ ൅ ݀௝௦ (4.9)
The process would be continued till either ݀௜௦ ൌ 0 or ݀௝௦ ൌ 0 and in view of (4.4) - (4.9), we
have
݀௜ ൌ ∑ ݊௞ߙ௜௞ ൅ ݀௜௦
௦
௞ୀଵ ߙ௜௞ ൐ ߙ௜,௞ାଵ (4.10)
݀௝ ൌ ∑ ݊௞ߚ௝௞ ൅ ݀௝௦
௦
௞ୀଵ ߚ௝௞ ൐ ߚ௜,௞ାଵ (4.11)
݀௨௦ ൌ ݊௦ାଵߜ௨,௦ାଵ ൅ ݀௨,௦ାଵ ሺ݀௨,௦ାଵ ൏ ߜ௨,௦ାଵሻ (4.12)
where ‫ݑ‬ ൌ ݅ or ݆, ߜ ൌ ߙ or ߚ for ݅ or ݆ respectively. Also ݊௞, ߙ௜௞, ߚ௝௞ are positive integers, may be
selected according to the length of stock in order to minimize the trim.
Referring relation (4.10)-(4.12), we now define the set
‫ܤ‬ ൌ ሼܾ௜௝
௥
, ܾ௨
௦ାଵ
, ܾ௨
௦ାଶ
: ܾ௜௝
௥
ൌ ߙ௜௥݈௜ ൅ ߚ௝௥݈௝, ܾ௨
௦ାଵ
ൌ ߜ௨,௦ାଵ݈௨, ܾ௨
௦ାଶ
ൌ
݀௨,௦ାଵ݈௨, where ܾ௜௝
௥
, ܾ௨
௦ାଵ
, ܾ௨
௦ାଶ
൑ ܷ௠, ‫ݑ‬ ൌ ݅ ‫ݎ݋‬ ݆, ߜ ൌ ߙ ‫ݎ݋‬ ߚ according as ‫ݑ‬ ൌ
݅ ‫ݎ݋‬ ݆ repectively, ‫ݎ‬ ൌ 1, 2, … , ‫.ݏ‬ ݅, ݆ ൌ 0, 1, … , ܰሽ (4.13)

529
Define |c୧| ൌ max c୧୨ , where c ൌ a or b for fixed ݅ and arbitrary ݆
ܿ௜௝ ൑ ܷ௠ (4.14)
In view of relation (4.13), we now define
‫ܤ‬௞
௥
ൌ ሼܾ௜௝
௥
, ܾ௨
௦ାଵ
, ܾ௨
௦ାଶ
: 0 ൏ หܷ௞ െ ൫ܾ௜௝
௥ |ܾ௨
௦ାଵ|ܾ௨
௦ାଶ
൯ห ൏ ‫ݐ‬௦
ఒ
(4.15)
ሺߣ ൌ 1, 2ሻ, ‫ݎ‬ ൌ 1, 2, … , ‫ݏ‬
5. CUTTING PLAN
It has been noticed practically that with the preference of starting from the largest order
lengths to the smaller ones, the cutting process has been executed in general as the smaller order
lengths left behind can be adjusted easily amongst them and results in less trim loss (see Figure 8.1).
• Cutting of the largest order length ࢒࢖ from category-I
Referring relation (4.14), we consider |ܽ௣|. In view of ‫ܣ‬௞ [cf. relation (4.10)], there exist sets
‫ܣ‬௤, ‫ܣ‬௥, ‫ܣ‬௦, … ሺ1 ൑ ‫,ݍ‬ ‫,ݎ‬ ‫,ݏ‬ … ൑ ݉ሻ containing |ܽ௣| along with some other ܽ௜௝’s. Corresponding to
each set ‫ܣ‬௤, ‫ܣ‬௥, ‫ܣ‬௦, … respective fixed stock lengths ܷ௤, ܷ௥, ܷ௦, … have been assigned. We select the
combination ܽ௣௤ corresponding to the smallest stock length ܷ௣ and focus our attention on it for the
first step of cutting.
Let หܽ௣ห ൌ ܽ௣௤(say) for ‫ݍ‬ ‫א‬ ‫ܤ‬ሺ‫݌‬ሻ~௣ where ܽ௣௤ ൌ ߙ௣݈௣ ൅ ߚ௤݈௤
satisfying the condition:
ௗ೛
ௗ೜
ൌ
ఈ೛ൈ௥೛೜
ఉ೜ൈ௥೛೜
(5.1)
In view of (5.1), it may be noted that by cutting ‫ݎ‬௣௤ bars of stock length ܷ௣, total number of
required pieces of order lengths ݈௣ and ݈௤ are cut.
Define
‫ݐ‬௣
ଵ
ൌ ‫ݎ‬௣௤൫ܷ௣ െ หܽ௣ห൯ ൑ ‫ݎ‬௣௤‫ݐ‬௦
ఒ
ሺߣ ൌ 1, 2ሻ (5.2)
• Cutting of other stock lengths from the set ࡭࢘
For ݅, ݆ ് ‫,݌‬ ‫,ݍ‬ we next consider the largest order length ݈௨ (say) contained in ‫ܣ‬௥ and consider
|ܽ௨| for ܽ௨௩ ‫א‬ ‫ܣ‬௞ corresponding to the stock length ܷ௨ satisfying the condition:
݀௨
݀௩
ൌ
ߙ௨ ൈ ‫ݎ‬௨௩
ߚ௩ ൈ ‫ݎ‬௦௩
for some ‫ݒ‬ ‫א‬ ‫ܤ‬ሺ‫݌‬ሻ~ሼ௣,௤,௨ሽ
Referring relation (5.1), it is clear that by cutting ‫ݎ‬௨௩ bars of the stock length ܷ௨, total
number of required pieces of order lengths ݈௨ and ݈௩ have been cut. Define
‫ݐ‬௨
ଵ
ൌ ‫ݎ‬௨ ሺܷ௨ െ |ܽ௨|ሻ ൑ ‫ݎ‬௨‫ݐ‬௦
ఒ
ሺߣ ൌ 1, 2ሻ (5.3)

530
Proceeding this way, for ݅, ݆ ് ‫,݌‬ ‫,ݍ‬ ‫,ݑ‬ ‫,ݒ‬ we consider the next largest order length out of the
remaining once and applying the same technique as before, the trim loss with respect to
corresponding stock lengths ܷ௝’s has been computed. The process is continued till all order lengths
belonging to category I are totally exhausted.
ܶଵ ൌ ∑ ‫ݐ‬௣
ଵ
௣‫א‬஻ሺ௡ሻ (5.4)
If this cutting process covers all the order lengths ݈ଵ, ݈ଶ, … , ݈௡, then STOP.
• Cutting of the largest order length ࢒ࡺ from Category-II
Referring definition of ‫ܤ‬௞
௥
[cf. relation (4.15)], we first set ‫ݎ‬ ൌ 1 and consider ܾே௝
ଵ
for fixed
ܰ and arbitrary ݆ and select ܾே௝
ଵ
as follows:
|ܾே| ൌ max
௝‫א‬஻ሺ௡ሻ~ಿ
ܾே௝
ଵ
Such that |ܾே| ൑ ܷ௝ for some ݆ ‫א‬ ‫ܤ‬ሺ݉ሻ.
Now, corresponding to |ܾே|, there exists sets ‫ܤ‬௤
ଵ
, ‫ܤ‬௥
ଵ
, ‫ܤ‬௦
ଵ
, … associated with the stock lengths
ܷ௤, ܷ௥, ܷ௦, … respectively containing |ܾே|. We select the set ‫ܤ‬௤
ଵ
corresponding to the smallest stock
length ܷ௤. In view of the relation (4.15), we have
ܾே௝
ଵ
ൌ ߙேଵ݈ே ൅ ߚ௝ଵ݈௝ for ݆ ‫א‬ ‫ܤ‬ሺܰሻ~ே.
It is clear from relations (4.10) and (4.11), that by cutting ݊ଵ bars of stock length ܷ௤, we cut
݊ଵ. ߙேଵ pieces of order length ݈ே and ݊ଵ. ߚ௝ଵ pieces of order length ݈௝. Our aim is to finish cutting of
only two order lengths first ݈ே and ݈௝ (fixed) at a time. Following cases may arise:
Case1. Either ݀ே െ ݊ଵ. ߙேଵ ൏ ߙேଵ or ݀௝ െ ݊ଵߚ௝ଵ ൏ ߚ௝ଵ or both the inequalities hold together.
Case 2. Either ݀ே ൌ ݊ଵ. ߙேଵ or ݀௝ ൌ ݊ଵߚ௝ଵ.
5.1 Remark
Here two cases will not hold together because in that case ݈ே and ݈௝ will belong to Category-I.
We first deal with the case 1. In view of the relation (4.13), we next consider
ܾேೕ
ଶ
ൌ ߙேଶ݈ே ൅ ߚ௝ଶ݈ଶ (݆ fixed as given by (4.13) )
Now, corresponding to ܾேೕ
ଶ
, there exist sets ‫ܤ‬௤
ଶ
, ‫ܤ‬௥
ଶ
, ‫ܤ‬௦
ଶ
, … containing it. The sets ‫ܤ‬௤
ଶ
, ‫ܤ‬௥
ଶ
, ‫ܤ‬௦
ଶ
, …
are associated with the stock length ܷ௤, ܷ௥, ܷ௦, … respectively. We select the set ‫ܤ‬௥
ଶ
(say)
corresponding to the smallest stock length ܷ௥. It is clear from relations (4.10) and (4.11) that by
cutting ݊ଶ bars of stock length ܷ௥, we cut ݊ଶ. ߙேଶ more pieces of order length ݈ே and ݊ଶ. ߚ௝ଶ more
pieces of order length ݈௝. We continue the process till either ݀ே ൌ ∑ ݊௞ߙ௜௞
௦
௞ୀଵ or ݀௝ ൌ ∑ ݊௞ߚ௝௞
௦
௞ୀଵ .
Let if possible ݀ே ൌ ∑ ݊௞ߙ௜௞
௦
௞ୀଵ holds, then ݀௝ would be of the form
݀௝ ൌ ∑ ݊௞ߚ௝௞
௦
௞ୀଵ ൅ ݀௝௦
where we express ݀௝௦ ൌ ݊௦ାଵߚ௝,௦ାଵ ൅ ݀௝,௦ାଵ ሺ݀௝,௦ାଵ ൏ ߚ௝,௦ାଵሻ.

531
Referring the relation (4.13), we now consider
ܾ௝
௦ାଵ
ൌ ߚ௝,௦ାଵ݈௝
Now, corresponding to ܾ௝
௦ାଵ
, there exists sets ‫ܤ‬௤
௦ାଵ
, ‫ܤ‬௥
௦ାଵ
, ‫ܤ‬௦
௦ାଵ
, … containing it. The sets
‫ܤ‬௤
௦ାଵ
, ‫ܤ‬௥
௦ାଵ
, ‫ܤ‬௦
௦ାଵ
, … associated with the stock lengths ܷ௤, ܷ௥, ܷ௦, … respectively. We select the set
‫ܤ‬௦
௦ାଵ
corresponding to the smallest stock length ܷ௦(say). It is clear from relations (4.10) and (4.11)
that by cutting of stock length ܷ௦, we cut ݊௦ାଵ.݀௝,௦ାଵ pieces of order length ݈௝. Now ݀௝,௦ାଵ pieces of
order length ݈௝ are left to cut out of ݀௝. We now consider ܷ௞ െ ݀௝,௦ାଵ݈௝ for all ݇ ൌ 1, 2, … , ݉ and
select the minimum difference corresponding to the stock length ܷ௧(say), all pieces of order length ݈௝
have been cut.
5.2 Remark
At this last step of cutting ܷ௧ െ ݀௝,௦ାଵ. ݈௝ may exceed the sustainable trim ‫ݐ‬௦.
We now compute the trim loss corresponding to the order lengths ݈ே and ݈௝ belonging to the
Category-II.
‫ݐ‬௥భ
ଵ
ൌ ቀܷ௤ െ ܾேೕ
ଵ
ቁ݊ଵ ൅ ቀܷ௥ െ ܾேೕ
ଶ
ቁ ݊ଶ ൅ ‫ڮ‬ ൌ ෍ ቀܷ௟ െ ܾேೕ
௥
ቁ݊௥ ൅ ሺܷ௧ െ ݀௝,௦ାଵ݈௝ሻ
௦
௥ୀଵ
Order lengths ݈ே and ݈௝ belonging to category-II have been cut completely. Remaining order
lengths we again arrange in increasing order ݈ଵ, ݈ଶ, … , ݈ெ (say). We first consider
|ܾெ| ൌ max
௝‫א‬஻ሺெሻ~ಾ
ܾெೕ
ଵ
݆ ‫א‬ ‫ܤ‬ሺ‫ܯ‬ሻ~ெ
such that |ܾெ| ൑ ܷ௝ for some ݆ ‫א‬ ‫ܤ‬ሺ݉ሻ.
Proceeding in a similar manner, we get
‫ݐ‬௥మ
′
ൌ ∑ ቀܷ௟ െ ܾெೕ
௥
ቁ ݊௥
௧
௥ୀଵ ൅ ൫ܷ௧ െ ݀௝,௦ାଵ݈௝൯ ݈ ‫א‬ ‫ܤ‬ሺ݉ሻ
We continue the process till all order lengths are exhausted and get
ܶଶ ൌ ‫ݐ‬௥భ
′
൅ ‫ݐ‬௥మ
′
൅ ‫ڮ‬
Finally, we get total trim
ܶ ൌ ܶଵ ൅ ܶଶ.
The percentages of trim lose with respect to ‫ݐ‬௦
ଵ
and ‫ݐ‬௦
ଶ
have been computed in accordance with
the total stock length used.

532
6. ALGORITHM

533

534
7. HAT FUNCTIONS AND THEIR APPLICATIONS IN 1D-CSP
Piecewise linear approximation may not have the practical significance in comparison to
cubic spline or even higher order approximation. But it shows most of the essential features of
piecewise polynomial approximation in a simple and easily understandable setting.
Consider the closed interval [a, b] on the real line and define a partition P:
ܲ: ܽ ൌ ߬ଵ ൏ ߬ଶ … , ൏ ߬௡ ൌ ܾ
Let $ଶdenote the collection or linear space of all continous broken lines on [a, b] with breaks at
߬ଶ, … , ߬௡ିଵ (see figure 7.1). We define the basis for $ଶ (cf. [3])
Let ߬଴: ൌ ߬ଵ , ߬௡ାଵ : ൌ ߬௡ and set
‫ܪ‬௜ሺ‫ݔ‬ሻ ൌ ቐ
ሺ‫ݔ‬ െ ߬௜ିଵሻ ሺ߬௜ െ ߬௜ିଵ ሻ⁄ , ߬௜ିଵ ൏ ‫ݔ‬ ൑ ߬௜
ሺ߬௜ାଵ – ‫ݔ‬ሻ/ሺ߬௜ାଵ െ ߬௜ሻ, ߬௜ ൑ ‫ݔ‬ ൏ ߬௜ାଵ
0 , otherwise
(7.1)
Figure 7.1 Hat Functions
It may be checked easily that the set S=ሼ‫ܪ‬௜ሺ‫ݔ‬ሻሽ௜ୀଵ
௡
is linearly independent. If we denote the
broken line interpolation to g by ‫ܫ‬ଶ݃, then it is given by
‫ܫ‬ଶ݃ ൌ ∑ ݃ሺ߬௜ሻ‫ܪ‬௜
௡
௜ୀଵ (7.2)
which satisfy the interpolatory conditions
‫ܫ‬ଶ݃ሺ߬௜ሻ ൌ ݃ሺ߬௜ሻ ݅ ൌ 1, … . . , ݊
7.1 Applications
For the given data, the average order lengths ߬௝ have to be computed say ߬଴ is the shortest and
߬௡ is the longest average order length. Consider the closed interval ሾ߬଴, ߬௡ሿ on ܴ and define the
partition ܲ as follows:
ܲ: ܽ ൌ ߬଴ ൏ ߬ଵ … . . ൏ ߬௡ ൌ ܾ (7.3)
Corresponding to each ߬௜ ሺ݅ ൌ 0, … . ݊ሻ, the total trim loss ሺ‫ݐ‬௜ ‫ݕܽݏ‬ሻ ‫ݐ‬௜ ൌ ݃ሺ߬௜ሻ by using the
algorithm described in section 6 has to be computed.

535
Similarly, corresponding to the sustainable trims of order one ‫ݐ‬௦
ଵ
and two ‫ݐ‬௦
ଶ
, we define a
partition and compute the broken line interpolation ‫ܫ‬ଶ݃ ߬߳ ሾ߬଴, ߬௡ሿ.
Since ‫ݐ‬௞′‫ݏ‬ are known, the basis functions (Hat functions) ‫ܪ‬௜′‫ݏ‬ as described in (7.1) can be
computed for the partition P in (7.3). Referring relation (7.2), we can write the broken line
interpolant ‫ܫ‬ଶ݃ for the approximate trim loss as follows:
‫ܫ‬ଶ݃ ൌ ෍ ݃ሺ߬௜ሻ‫ܪ‬௜
௡
௜ୀଵ
which gives the information of approximate trim loss at any arbitrary point ߬߳ ሾ߬଴, ߬௡ሿ.
7.1 Remark
Broken lines are neither very smooth nor very efficient approximation. Both for a smoother
and more efficient approximation one has to go to piecewise polynomial approximation with higher
order pieces. The most popular choice continues to be a piecewise cubic approximating function.
Since, it is the first step of splines to enter into a field of cutting-stock problem (CSP), authors have
started with the simplest formulation of approximate function for the trim loss. Higher order
approximation would be constructed in accordance with the demand of this method in the practical
field.
8. ILLUSTRATIVE EXAMPLE
Tables of data
Order lengths and required number of pieces
S.No.
Order lengths
(in cm.)
Required no. of
pieces
S.No. Order lengths
(in cm.)
Required no. of
pieces
1. 801 03 6. 498 16
2.
748 24 7. 492 39
3.
733 46 8. 471 21
4.
641 23 9. 327 40
5.
548 39 10. 303 32
Table 8.1
Available stock lengths
S.No.
Stock lengths (in cm.) S.No. Stock lengths (in cm.)
1.
2110 4. 3883
2. 2210 5. 4177
3. 3120 6. 4239
Table 8.2

536
Figure 8.1 Cutting Plan
Figure 8.2 Screen shot of the programming
Figure 8.3 Screen shot of the programming

537
Computations as discussed in sections 3 and 5
First order sustainable trim ሺ࢚࢙
૚
ሻ = 66.3591,
Second order sustainable trim ሺ࢚࢙
૛
ሻ = 68.1928
S. No. Sustainable Trims Total Trim in %
1. 66.3591 0.3461
2. 66.8176 0.3461
3. 67.2756 0.1225
4. 67.7342 0.1225
5. 68.1928 0.1225
Table 8.15
Partition ܲ = {߬ଵሺൌ 66.3591ሻ ൏ ߬ଶሺൌ 66.8176ሻ ൏ τଷሺൌ 67.2756ሻ ൏ τସሺൌ 67.7342ሻ ൏
τହሺൌ 68.1928ሻ}
Figure 8.4 Linear splines for total trim loss corresponding to various sustainable trims
9. CONCLUSION
It is interesting to note that the impact of sustainable trim on the total trim plays key role in
1D-CSP. Referring Figure 8.4, it is apparent that corresponding to the second order sustainable trim
ሺ‫ݐ‬௦
ଶ
ሻ which is greater than the sustainable trim of order one ሺ‫ݐ‬௦
ଵ
ሻ, the total trim loss is minimal
corresponding to ሺ‫ݐ‬௦
ଶ
ሻ in the interval ሾ‫ݐ‬௦
ଵ
, ‫ݐ‬௦
ଶ
ሿ. It is suggested that for a given data, and given stock
lengths, after computing ሺ‫ݐ‬௦
ଵ
ሻ and ሺ‫ݐ‬௦
ଶ
ሻ, the minimal trim loss may be predicated locally as per the
proposed cutting plan.
REFERENCES
1. Belov, G., Scheithauer, G. (2006), A branch-and-cut-and-price algorithm for one-dimensional
stock cutting and two-dimensional two-stage cutting, European Journal of Operational
Research, 171(1), 85-106.
2. Belov, G., Scheithauer, G. (2002), A cutting plane algorithm for the one-dimensional cutting
stock problem with multiple stock lengths, European Journal of Operational Research,
141(2), 274-294.

538
3. Boor, Carl de (1978), A Practical Guide to Splines, Springer Verlag.
4. Dikili, A.C., Barlas, B. (2011), A generalized approach to the Solution of one-dimensional
stock-cutting problem for small shipyards, Journal of Marine Science and Technology, 19(4),
368-376.
5. Dyckhoff, H. (1981), A new linear programming approach to the cutting-stock problem,
Operations Research, 29, 1092-1104.
6. Gilmore, P.C., Gomory, R.E. (1965), Multistage cutting stock problems of two and more
dimensions. INFORMS.Operations Research 13(1), 94-120.
7. Gilmore, P., Gomory, R. (1963), A linear programming approach to the cutting stock
problem, part-II, Operations Research 11, 863-888.
8. Gilmore, P., Gomory, R. (1961), A linear programming approach to the cutting stock
problem, Operations Research 9(6), 848-859.
9. Gradisar,M., Resinovic, G., Kljajic, M. (2002), Evaluation of algorithms of one-dimensional
cutting, Computer & Operations Research 29, 1207-1220.
10. Gradisar, M., Jesenko, J., Resinovic, G. (1997), Optimization of roll cutting in clothing
industry, Computers Ops. Res., 24(10), 945-953.
11. Powar, P.L., Jain, V., Saraf, M., Vishwakarama, R. (2013), One-Dimensional Cutting Stock
Problem (1D-CSP) with second order sustainable trim: a comparative study with first order
sustainable trim, International Journal of Computer Engineering and Technology (IJCET),
ISSN 0976 – 6367(Print) ISSN 0976 – 6375(Online), 4(4), 136-149.
Problem (1D-CSP) with First Order Sustainable Trim: A Practical Approach, International
Journal of Computer Science Engineering and Information Technology Research (IJCSEITR)
ISSN 2249-6831, TJPRC Pvt. Ltd., 3(3), 227-240.
Problem (1D-CSP): A New Approach, International Journal of Advanced Computing, Recent
Science Publication, 46(2), 1223-1228.
14. Stadtler, H. (1990), A one-dimensional cutting problem in the aluminum industry and its
solution, European Journal of Operational Research, 44, 209-223.
15. Tilanus, C., Gerhardt, C. (1976), An application of cutting stock in the steel industry,
Operational Research, 669-675.
16. Tokuyama, H., and Ueno, N. (1981), The cutting stock problems in the iron and steel
industries in J.P. Brans (ed.), Operational Research, North- Holland, 809-823.
17. Vance, P. (1998), Branch-and-price algorithms for the one-dimensional cutting stock
problem, Computational Optimization and Applications, 9(3), 211-228.
18. Vanderbeck, F. (2000), On Dantzig-Wolfe decomposition in integer programming and ways
to perform branching in a branch-and-price algorithm, Operation Research, 48, 111-128.
19. K. V. Sujatha and S. Meenakshi Sundaram, “Regression, Theil’s and Mlp Forecasting
Modelsof Stock Index”, International Journal of Computer Engineering & Technology
(IJCET),Vol 1, Iss 1, 2010, pp. 82 - 91, ISSN Print: 0976 – 6367, ISSN Online: 0976 – 6375.
20. Rajesh V. Argiddi and Sulabha S. Apte, “A Study of Association Rule Mining in Fragmented
Item-Sets for Prediction of Transactions Outcome in Stock Trading Systems”, International
Journal of Computer Engineering & Technology (IJCET), Vol 3, Iss 2, 2012, pp. 478 - 486,
ISSN Print: 0976 – 6367, ISSN Online: 0976 – 6375.

One dimensional cutting stock problem 1-d-csp_ a study on data dependent tri

More Related Content

What's hot (19)

Viewers also liked (9)

Similar to One dimensional cutting stock problem 1-d-csp_ a study on data dependent tri (20)

More from IAEME Publication (20)

Recently uploaded (20)

One dimensional cutting stock problem 1-d-csp_ a study on data dependent tri