alexa
Reach Us +44-1474-556909
A Solid Transportation Problem with Partial Non-Linear Transportation Cost | OMICS International
ISSN: 2168-9679
Journal of Applied & Computational Mathematics
Make the best use of Scientific Research and information from our 700+ peer reviewed, Open Access Journals that operates with the help of 50,000+ Editorial Board Members and esteemed reviewers and 1000+ Scientific associations in Medical, Clinical, Pharmaceutical, Engineering, Technology and Management Fields.
Meet Inspiring Speakers and Experts at our 3000+ Global Conferenceseries Events with over 600+ Conferences, 1200+ Symposiums and 1200+ Workshops on Medical, Pharma, Engineering, Science, Technology and Business
All submissions of the EM system will be redirected to Online Manuscript Submission System. Authors are requested to submit articles directly to Online Manuscript Submission System of respective journal.

A Solid Transportation Problem with Partial Non-Linear Transportation Cost

Anupam Ojha*, Shyamal Kumar Mondal and Manoranjan Maiti

Department of Applied Mathematics with Oceanology and Computer Programming, Vidyasagar University, Midnapore - 721102, India

*Corresponding Author:
Anupam Ojha
Department of Applied Mathematics with
Oceanology and Computer Programming
Vidyasagar University, Midnapore–721102, India
Tel: 3222 276 554
E-mail: [email protected]

Received August 17, 2013; Accepted January 02, 2014; Published January 09, 2014

Citation: Ojha A, Mondal SK, Maiti M (2014) A Solid Transportation Problem with Partial Non-Linear Transportation Cost. J Appl Computat Math 3: 150. doi: 10.4172/2168-9679.1000150

Copyright: © 2014 Ojha A, et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Visit for more related articles at Journal of Applied & Computational Mathematics

Abstract

This paper deals with a capacitated, multi-objective and solid transportation problem with imprecise nature of resources, demands, capacity of conveyance and cost. Here transportation cost is inversely varying with the quantity-to be transported from source to destinations in addition with a fixed per unit cost and a small vehicle cost. The transportation problem has been formulated as a constrained fuzzy non-linear programming problem. Next it is transformed into an equivalent crisp multi-objective problem using fuzzy interval approximation and then solved by Interactive Fuzzy Programming Technique (IFPT) and Generalized Reduced Gradient (GRG) method. An illustrative numerical example is demonstrated to find the optimal solution of the proposed model.

Keywords

Transportation models; Fuzzy numbers; Interval approximation; Interactive fuzzy programming technique

Introduction

The classical transportation problem (Hitchcock transportation problem) is one of the sub-classes of non-linear programming problem in which all the constraints are of equality type. In many industrial problem, a homogeneous product is delivered from an origin to a destination by means of different modes of transport called conveyances, such as trucks, cargo flights, goods trains, ships, etc. In general, the real life problems are modeled with multi-objective functions which are measured in different respects and they are non-commensurable and conflicting in nature. Furthermore, it is frequently difficult for the decision maker to combine the objective functions in one overall utility function. In a Solid Transportation Problem (STP) more than one objective is normally considered. In many practical situations, it is realistic to assume that the amount which can be sent on any particular route is restricted by the capacity of that route. Moreover, Appa [1] discussed about the different variations in transportation problem.

It is often difficult to estimate the accurate values of transportation cost, delivery time, quantity of goods delivered, demands, availabilities, the capacity of different modes of trans-port between origins to destinations, etc. Depending upon different aspects, these fluctuate due to uncertainty in judgment, lack of evidence, in sufficient information, etc. i.e., it is not possible to get relevant precise data, which are assumed by several researchers Shell [2]. So, a transportation model become more realistic if these parameters are assumed to be flexible/imprecise in nature i.e., uncertain in non-stochastic sense and may be represented by fuzzy numbers. (Here, for the first time a solid transportation problem is considered with fuzzy parameters like, Shell [2], Jimenez and Verdegay [3].

Based on Das et al. [4], the interval number transportation problems were converted into deterministic multi-objective problems. Grzegorzewski [5] approximated the fuzzy number to its nearest interval. Omar and Samir [6] and Chanas and Kuchta [7] discussed the solution algorithm for solving the transportation problem in fuzz environment. Sakawa and Yano [8] proposed an interactive fuzzy decision making method using linear and non-linear membership functions to solve the multi-objective linear programming problem. Gao and Liu [7,9] presented two-phase fuzzy algorithms. Shaocheng [10] discussed about the interval number linear programming. Verma et al. [11], Bit at al. [12,13], Jimenez and Verdegay [14], Li and Lai [15] and Waiel [16] presented the fuzzy compromise programming approach to multi-objective transportation problem.

In this paper, a capacitated-multi-objective, solid transportation problem is formulated in fuzzy environment with non-linear varying transportation charge and an extra cost for transporting the amount to an interior place through small vehicles (like rickshaw, auto etc.). In the non-linear varying transportation charge, one part is linearly and another nonlinearly proportional to the transported amount. Here, the non-linear cost increases with the increase of transported amount but the rate of increase decreases. The fuzzy quantities and parameters are replaced by equivalent nearest interval numbers and thus a fuzzy multiobjective, capacitated and solid transportation problem is transformed to corresponding crisp multi-objective transportation problems. Membership function is formulated for each objective function from their individual minimum and maximum. These membership functions may be of different types and it may depend upon the decision maker’s (DM) choice/past data, if available. The main contribution of the paper is the mathematical formation of the above mentioned innovative transportation problem in fuzzy environment and its solution. Here for the first time, here man-machine interaction has been introduced in the transportation system. In this way, multi-objective transportation problem is solved using IFPT through GRG. This model is illustrated with an example. In particular, results of some specific transportation models are presented.

Fuzzy Number and its Interval Approximation

Definition 1: Fuzzy Number [17]:

If X is a collection of objects denoted generically by x then a fuzzy set à in X is a set of ordered pairs:

image

image is called the membership function of x in imagewhich maps X to the membership space [0,1].

(i) A is normal, i.e.the supremum of

(ii) A is fuzzy convex, i.e. μA(x) is 1 ∀x∈S and

(iii) A is fuzzy convex, i.e.

image

Definition 2: α-Cut of a fuzzy number [17] (Figure 1):

computational-mathematics-general-fuzzy-number

Figure 1: α-cut of a general fuzzy number.

A α-cut of a fuzzy number A is defined as crisp set

Aα={x : μ ˜(x) ≥ α, x ∈ X} where α ∈ [0, 1]

Definition 3: Interval Approximation [17]:

Aα is a non-empty bounded closed interval contained in X and it can be denoted by Aα=[AL(α), AR(α)]. AL(α) and AR(α) are the lower and upper bounds of the closed intervalrespectively. A fuzzy number A with α1, α2-cut Aα1=[AL(α1), AR(α1)],

Aα2=[AL2), AR2)] and if α2 ≥ α1, then AL2) ≥ AL1) and AR1) ≥ AR2).

Definition 4: Nearest interval approximation of a fuzzy number [18]:

Here a fuzzy number is approximated by a corresponding crisp interval. Suppose A and B, two fuzzy numbers with α-cuts are [AL(α), AR(α)] and [BL(α), BR(α)] respectively. Then the distance between imageand image is

image

Given A is a fuzzy number. We find a closed interval image which is nearest to imagewith respect to metric d. It can be done since each interval is also a fuzzy number with constant α-cut for all α € [0, 1]. Hence (Cd(A))α=[CL, CR]. Now we have to minimize

image

with respect to CL and CR. In order to minimize image, it is sufficient to minimize the function image. The first partial derivatives are

image

Solving,

image

we get,

imageand image

Again since, image

and image

So D(CL, CR) i.e. imageis global minimum. Therefore the interval

imageis nearest interval approximation of fuzzy number image with respect to metric d.

Let imagebe a fuzzy number. The α-level interval of A is defined as image.

When imageis a trapezoidal fuzzy numbers then AL(α)=a1+α(a2−a1) and AR(α)=a3−α(a4−a3), 0<α ≤ 1.

By nearest interval approximation method lower and upper limits of the interval are respectively

image

and image

Therefore, the nearest crisp interval number considering imageas a trapezoidal fuzzy number is [(a1+a2)/2, (a3+a4)/2].

Interval Analysis

We consider an interval A=[aL, aR]={a:aL ≤ a ≤ aR, a ∈ R}, where aL and aR are the left and right limits of A respectively. The interval A is also denoted by its centre and width as:

image

Where aC=(aR+aL)/2 and aW=(aR−aL)/2 are the center and half-width of A respectively.

Definition 5 [18]: The order relation, ≤RC between A=[aL, aR] and B=[bL, bR] is defined as

image

Where aC=(aR+aL)/2 and aW=(aR−aL)/2 are the center and half-width of A respectively.

Definition 5 [18]: The order relation, ≤RC between A=[aL, aR] and B=[bL, bR] is defined as

image

The order relation ≤RC represents the decision maker’s preference for the alternative with the lower minimum value, i.e., if A ≤RC B, then A is preferred to B for the minimization problem.

Theorem: The order relation ≤RC satisfied transitive law.

Proof: Let A=[aL, aR], B=[bL, bR] and C=[cL, cR] such that

image

Then subtracting we have aL ≤ bL ≤ cL and

A ≤RC B & B ≤RC C ⇒ A ≤RC C

image

If possible image

image

Moreover aL=cL and aR=cR

image

image

which is the contradiction. Therefore image. Thus the theorem proved.

Formulation of the STP in Fuzzy Environments

The solid transportation problem (STP) stated by Shell [2] is a generalization of classical transportation problem. Here, we consider m origins(or sources) Oi (i=1,2,....m) , n desti-nations(i.e. demands) Dj (j=1,2,....n) and K conveyances Ek (k=1,2,....K). Let image be the fuzzy amount of a homogeneous product available at i-th origin, imagebe the fuzzy demand at j-th destination and imagerepresents the fuzzy amount of product which can be carried by k-th conveyance. The fuzzy penalty image is associated with transportation cost (for p=1), distance, time etc, (for p=2,3..P) of one unit of a product to transport from i-th source to j-th destination by means of the k-th conveyance for p-th criterions. image be the inversely varying cost for different quantity to transport from i-th source to j-th destination by means of the k-th conveyance. The penalty may represent transportation cost, delivery time, quantity of goods delivered etc. The vehicle carrying cost F (xijk) for the quantity xijk from i-th source Oi to j-th destination Dj via k-th conveyance is defined as:

image

m=[xijk/vc], vc=vechicle capacity and v=vehicle cost.

Let rijk be the capacity restrictions on route (i,j) by means of k-th conveyance. Therefore, the fuzzy-capacitated -constrained multicriteria solid transportation problem represented as:

image(1)

image(2)

0 <γijk< 1

subject to the constraints

image(3)

image(4)

image(5)

image(6)

Where imageare fuzzy numbers which may be represented by triangular, trapezoidal or other fuzzy numbers.

Reduced Crisp Model

Following ($4) fuzzy numbers imageare approximated to

image respectively. Then the earlier transportation model takes the following form:

image(7)

image(8)

image(9)

image(10)

image(11)

image(12)

subject to the constraints

image(13)

image(14)

image(15)

image(16)

The constraints (13),(14),(15) and (16) can be written in the following form:

image(17)

image(18)

image(19)

image(20)

Solution Procedure: Interactive Fuzzy Programming Technique

Step-1: Optimize (minimized & maximized) each objective function individually at a time ignoring the others. Let, image and image be the solutions for the objective functions image and image(p=1,2,3,...,P) respectively [8].

Step-2: Determine the corresponding values for every objective at each solution derived and form a pay-off (Table 1).

x ZR1 ZC1 ZR2 ZC2 - - ZCP
xR1 ZR1(xR1) ZC1 (xR1) ZR2(xR1) ZC2 (xR1) - - ZCP (xR1)
xC1 ZR1(xC1) ZC1(xC1) ZR2(xC1) ZC2(xC1) - - ZCP (xC1)
xR2 ZR1(xR2) ZC1(xR2) ZR2(xR2) ZC2(xR2) - - ZCP(xR2)
xC2 ZR1(xC2) ZC1(xC2) ZR2(xC2) ZC2(xC2) - - ZCP (xC2)
- - - - - -   - -
- - - - - -   - -
xRP ZR1(xRP) ZC1(xRP) ZR2(xRP) ZC2(xRP) - - ZCP(xRP)
xCP ZR1(xCP) ZC1(xCP) ZR2(xCP) ZC2(xCP) - - ZCP (xCP )

Table 1: Pay-off Matrix for the objective functions.

Also consider image and imagebe the upper and lower bounds of imagefunctions (i=1,2,3,....,P) from Table 1 respectively.

Step-3: Let image and image be the membership functions of imageand imagerespectively. The decision maker can select the above membership functions in a subjective manner from among the following four types of functions: linear, exponential, hyperbolic and hyperbolic inverse functions.

Different types of membership functions: Let ZU and ZL be the upper and lower bounds of the function Z(x) respectively. For the function Z(x), the corresponding membership functions are defined as follows:

(i) (Type-I) Linear membership function:

μZ (x)=[ZU−Z(x)]/[ZU−ZL](21)

(ii) (Type-II) Exponential membership function:

μZ (x)=a[1−exp{−b([ZU−Z(x)]/[ZU−ZL])}](22)

image

(iii) (Type-III) Hyperbolic membership function:

image(23)

where c=tanh-11.

(iv) (Type-IV) Hyperbolic inverse membership function:

image(24)

whereimage

Step-4: After determining the membership functions for each of the objective functions, we adopt the maximizing decision proposed by Bellman and Zadeh [19], the resulting problem to be solved is:

Max (min of all μ(x))

subject to x ∈ S where S is the feasible region of problem.

By introducing an auxiliary variable β, the above problem can be transformed into the following equivalent conventional problem.

Maximize β,(25)

image(26)

image(27)

where 0 ≤ β ≤ 1 and x ∈ S (the set of all feasible solutions)

However, with the four types of membership functions given by (21),(22),(23) and (24), the resulting problem is a non-linear programming problem. The constraint like μz(x) ≥ β takes the following form:

image(28)

Where imagefor Types -I, II, III and IV respectively. The solution of this final problem gives β=β* and the corresponding x=x∗ is the required optimum solution.

Numerical Experiment

To illustrated proposed model numerically, we a set of input data for two origins and destination, three different conveyance and criteria (Table 2).

image

image

p = 1 p = 2 p = 3  
(1, 3, 5, 7) (2, 5, 6,9) (6, 9, 10,12) (1, 2, 6, 7)
(3, 7, 9, 11) (7, 9, 12,15) (6, 9, 11,14) (4, 5, 7, 12)
(3, 6, 9, 11) (6, 9, 13,17) (5, 9, 11,13) (3, 7, 9, 12)
(2, 5, 7, 12) (5, 8, 11,15) (5, 7, 9, 12) (5, 9, 11, 13)
(2, 3, 5, 9) (1, 5, 9, 10) (3, 7, 10,13) (3, 5, 9, 12)
(4, 6, 8, 12) (4, 7, 8, 14) (7, 9, 11,15) (6,10,12,15)
(3, 7, 9, 12) (2, 7, 10,13) (6, 9, 13,15) (4, 6, 8, 10)
(5, 8, 10, 13) (3, 8, 12, 16) (4, 7, 9, 12) (6, 9, 11, 14)
(4, 5, 7, 9) (4, 8, 12, 14) (1, 3, 5,   7) (2, 4, 6, 10)
(5, 9, 12, 14) (3, 7, 11, 13) (3, 5, 8, 12) (3, 7, 10, 13)
(6, 8, 13, 15) (6, 9, 12, 15) (2, 5, 9, 10) (3, 5, 9, 13)
(3, 5, 9, 13) (5, 9, 12, 15) (5, 9, 13, 16) (4, 6, 10, 12)

Table 2: Input data for transportation costs.

v=12, Vc=5 and

Let the route capacities, rijk’s are

r111=16; r121=19; r211=14; r221=15; r112=17; r122=19; r212=16; r222=14;

r113 =16 ; r123=19 ; r213=14 ; r223=16;

γ111=0.3; γ121=0.7; γ211=0.45; γ221=0.75;

γ112=0.25; γ122=0.5; γ212=0.25; γ222=0.35;

γ113=0.75; γ123=0.8; γ213=0.95; γ223=0.85;

We now approximate the fuzzy numbers to their nearest intervals, then form the objective functions image, image, image, image, image, image find their individual solutions image, image, image, image, image, image, image minimizing image, image, image, image, image, and image separately.

The individual minimum and maximum of image, image, image, image, image, image are shown is pay-off table in Table 2. Now solving the problem (25), with all membership functions as linear, we get Table 3.

Obj.
Max. 817.548 1036.145 582.368 817.282 630.437 772.24
Min. 718.0256 936.0466 522 751 467.812 615.5

Table 3: Pay-off table.

Following IFPT, the auxiliary variable β is maximized and the optimal solutions are displayed in Tables 3 and 4 and corresponding decision variable are again if we take the membership functions μ1C, μ2C and μ3R as exponential and others linear, then the solution and the objective functions are shown in Tables 5 and 6, and corresponding decision variable are shown in Table 7.

β* ZC1* ZR1* ZC2* ZR2* ZC3* ZR3* Zav
0.426653 775.086 988.212 556.612 789.001 561.052 693.894 727.31

Table 4: Optimal auxiliary variable.

Conv-1 Conv-2 Conv-3
12.418 0 5.903 6.676 11 0
0 6.582 3.676 8.244 0 11.497

Table 5: The transported amounts.

β* ZC1* ZR1* ZC2* ZR2* ZC3* ZR3* Zav
0.25486E-06 804.934 1021.602 551.658 775.665 615.786 757.824 754.58

Table 6: Optimal auxiliary variable.

β* ZC1* ZR1* ZC2* ZR2* ZC3* ZR3* Zav
0.394097 766.902 976.196 570.267 793.062 550.606 698.779 725.97

Table 7: The transported amounts.

Particular cases

Case I: Multi-objective and Solid Transportation Problem (i.e., Non-capacitated)

In this case all routes can be used to transport any amount of products. For this situation we remove the route restrictions (rijk) from the formulation of above problem. For the reduced problem we get the following optimal solution using all the linear membership functions and corresponding decision variable are displayed in Tables 8 and 9.

β* ZC1* ZR1* ZC2* ZR2* ZC3* ZR3* Zav
0.394097 766.902 976.196 570.267 793.062 550.606 698.779 725.97

Table 8: Optimal auxiliary variable.

Conv-1 Conv-2 Conv-3
6.938 0 5.009 9.863 14.184 0.004
0 12.061 6.867 2.758 0 8.312

Table 9: The transported amounts.

Case II: Multi-objective and Capacitated Transportation Problem ( i.e., Non-solid)

In this case only one conveyance is used for transportation with some route restrictions. Since every conveyance has some capacity, sometimes one conveyance cannot transport all products. Due to this reason we take individual conveyance for transportation with route restrictions and all membership functions as linear, the first conveyance is not suitable for transportation: Table 10 and corresponding decision variable are for conveyance-2 & conveyance- 3 shown in Tables 11 and 12 respectively.

β* ZC1* ZR1* ZC2* ZR2* ZC3* ZR3* Zav
0.020707 824.98 981.033 485.622 688.26 570.573 741.198 715.28

Table 10: Non-solid and capacitated for conveyance-2.

13.01 18.989
14.489 13.510
14.265 17.73
13 15

Table 11: The transported amounts.

β* ZC1* ZR1* ZC2* ZR2* ZC3* ZR3* Zav
0.463242 701.649 893.878 576.515 776.265 426.945 603.929 663.197
0.11801 770.265 956.16 488.064 678.495 619.398 790.023 717.067

Table 12: Non-solid and capacitated for conveyance-2.

Case III: Transportation Problem not being solid and capacitated

This case is the combination of above two cases (Case I and II). Taking all the membership functions as linear, we find the following optimal solution for the individual conveyance-2 displayed in Tables 12 and 13) or conveyance 3 in Table 14. But if we take only conveyance-1, then feasible solution does not exist, and corresponding decision variable are for conveyance-2 shown in Tables 12 and 15.

3.245 28.754
24.254 3.745

Table 13: The transported amounts.

β* ZC1* ZR1* ZC2* ZR2* ZC3* ZR3* Zav
0.460702 699.197 890.871 576.90 776.717 426.618 603.33 662.273

Table 14: For not solid and not capacitated for conv-3.

14.717 17.283
12.76 15.24

Table 15: The transported amounts.

Conclusion

This paper proposes an interactive fuzzy programming approach to find the optimal com-promise solution for an innovative capacitated, multi-objective, solid transportation problem with fuzzy fixed charge, fuzzy partial varying transportation charge and vehicle cost. Here the transportation model is more realistic and flexible in nature. In the proposed problem the coefficient of objective function, resources, demands, and conveyances are taken as trapezoidal fuzzy number and approximated to corresponding nearest interval. Finally the solution procedures have been illustrated by an example. The present formulation and solution procedures can be applied to transportation problems with other and general fuzzy numbers.

References

Select your language of interest to view the total content in your interested language
Post your comment

Share This Article

Article Usage

  • Total views: 12750
  • [From(publication date):
    January-2014 - Jul 18, 2019]
  • Breakdown by view type
  • HTML page views : 8833
  • PDF downloads : 3917
Top