Scribd Logo
Upload

Welcome to Scribd!

  • 20005a78 920 PDF

    Uploaded by

    Amit
    14 views10 pages

    Original Title

    20005a78_920.pdf

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    Report this Document

    Copyright:

    © All Rights Reserved
    Download as PDF or read online from Scribd
    Flag for inappropriate content
    14 views10 pages

    20005a78 920 PDF

    Uploaded by

    Amit

    Copyright:

    © All Rights Reserved
    Download as PDF or read online from Scribd
    Flag for inappropriate content
    of 10
    A PROCEDURE FOR TIME MINIMIZATION TRANSPORTATION PROBLEM. by HL. BHATIA, Department of Mathematics, P.G.D.A.V. College, New Delhi 110 024 Kani Swaroor, Indian Institute of Public Administration, New Delhi and M. C. Purt, Department of Mathematics, Hans Raj College, Delhi University, Delhi 110 007 (Received 8 September 1975; after revision 14 July 1976) This paper develops a technique for minimizing time in a transportation problem. The procedure involves finite iterations and is based on moving from a basic feasible solution to another till the last solution is arrived at. ‘A numerical example illustrating the method is also included. INTRODUCTION In a time minimizing transportation problem, the time of transporting goods. from m origins to n destinations is minimized, satisfying certain conditions in respzct of availabilities at sources and requirements at the destinations. Thus, a time minimizing transportation problem is: Minimize Z=[Max ty,/x,,> 0] in subject to B xy=a,i=1,2,..m Problem (1) a oy E x= by fa an a xy 20 Here 14, is the time of transporting goods from the ith origin, where availability: is q, to the jth destination, where the requirement is 4, For any given feasible solution, X= [xy] satisfying (1), the time of transportation is the maximum of ty's among the cells in which there are positive allocations, i.e., corresponding to the solution X, the time of transportation is (Max tyj[x> 0]. We Vol. 8, No. 8 ‘TIME MINIMIZATION TRANSPORTATION PROBLEM 921 ‘The aim is to minimize this time of transportation. Such problems arise when it is required to transport perishable goods or, during war days, it is required to transport food and armaments in the shortest possible time and in so many other similar situations. Thus, the basic difference between the cost minimizing transportation problem and the present problem is that whereas the cost of transportation changes with variations in the quantity of the commodity, the time involved remains unchanged, irrespective of the quantities of the commodity involved in the occupied cells in the present problem. It is assumed in the given problem that (i) the carriers have sufficient capacity to carry goods from an origin to a destination in a single trip, and (ji) they start simultaneously from their respective origins. The time minimizing transportation problem has been studied by Hammer (1969), Garfinkel and Rao (1971) and Szware (1971). Hammer (1969) and Szware (1971) used lebelling techniques to solve the problem (1). Garfinkel and Rao (1971) solved the problem by introducing a sufficiently large cost M on certain routes. In the present study, the labelling technique has not been used and the costs associated are just 0 and 1 and do not involve costs like M, as was done by Garfinkel and Rao (1971). Thus, the solution method presented in this paper is quite simple from computational point of view. TaEorETICAL DEVELOPMENT Definitions (® A feasible solution—A set X = [x,] of non-negative variables satisfying (1) is called a feasible solution. (ii) A better feasible solution—Let X* = [xh] and X?= [xj] be two feasible solutions of problem (). Let M = [(,/)/x> 0], M? = [G/)/x4> 0 Ti= Max 1, T?= Max 1 seme . Gem “ P= GDtu=T Ge] P=[EAty=T GEM] A solution X? is said to be better than X’, if either (1) T? 0] is minimal is ras called an almost optimal solution, Optimal Solution A feasible solution X = [x,] of the Problem (1) is said to be the optimal solu- tion if there does not exist any better solution, i.e., for which the time of trans- Portation 7 is the least, and | 5 ay also has the minimum value, Wireo Theorem \—If there is a feasible solution to a set of simultaneous equations AX = b, X >0, then there is a basic feasible solution (refer Hadley 1962, Arts. 3.4, p. 80). Hadley (1962) proved this theorem by reducing the number of positive variables. inthe given feasible solution one by one till the columns of A associated with positive variables are linearly independent. In the proof of this theorem, the set of positive variables in the basic feasible solution is a subset of the positive variables in the given feasible solution. The values of the variables in the two sets, however, may be different. Theorem 2—-There exists a basic feasible solution which is optimal for Problem (1). Proor: Let [tag] represent a set of ty’s corresponding to those x,y’s that form a feasible solution. Let [#3] be a set of t,4’s corresponding to those positive %,’s which form the basic feasible solution derived from the feasible solution. Since [ta] is essentially a subset of [fag] the maximum of ty’s in [ty3] is less than or equal to the maximum of /,'s in [fag] It follows that if there exists a feasible solution of (I) with the time of transportation T, there also exists a basic feasible solution with the time of transportation 7’, such that T’T (the cells in which the time of transportation is greater than or equal to 7? are considered non-economical carrying no profit). C! = Profit yielded by solution X1= FF cy xif en oe = 35 x=N-p 5 eat where N= J a= 2 by < 4 C? = Profit yielded by solution X? = 4 by xy. nea As Xtis a better solution than X%, either T?< T! or T?=T%, p? N—pt, C?> Cl, When T?=T, pt < pi. C= LE cyxy = FE xy wyem oy eibiae Nap As N— pt> N—p!, C> Ch, It follows that X7 is not the optimal solution for the profit-maximization trans- portation problem. Therefore, there exists an adjacent basic feasible solution X* for the profit maximization problem better than X%, such that C’ > C', where C? is the profit yielded by solution X*. Define M*, T%, Q%, p* in the usual manner for the solution X*. It can be shown that 7? <7". If possible, let T? > 7". Now there may exist a possibility in which C* contains all the cells of X* with time 77 and the allocation same as in X*, Then 2 cox = =N—p Om Bh OO oa iaugs THAN P As Cl=N—p! and N—p'>N—p'—p! = Cl>C or O Ci>N—p'> N—p! = p 1 2=[GDita=T.GNEM) Define a profit matrix [cu] as in Theorem 2, ie., = [3 if ty 0, where Bra = [ere — (ur + P)/Cr, 8) & B], where B is the set of basic cells. Acell (r,s) with , > T, C= 0 € S, for if such a cell enters the basis at a positive level, the profit decreases, but our problem is a profit maximization problem. In case it enters at zero level, neither time nor profit changes, so that it is an alternative degenerate solution. Thus, such cells are not eligible for entry into the basis. For cells (r,s) with tw<.T, ¢e=1, An is positive if an only is u, + ¥, <0. A S=[O s,s) EB, ut, <0, Co=1] The procedure is bound to converge, for it involves movement from one basic feasible solution to another better basic feasible solution; the solutions are always finite in number. The process terminates when S= ®, Note~An imbalanced transportation problem, when Zax ¥ by can be solved by introducing either a dummy destination (when Sa, > _ with demand, ~ 2, or a dummy source (when Za, < Eb) with a capacity 3b, — Baw The time of transportation to dummis will be less than the ininimom 4, in the time matrix [f]. NUMERICAL ILLUSTRATION Let there be six producers O, supplying 15, 7, 45, 30, 12, and 16 units of goods respectively with seven consumers D, demanding 20, 13, 11, 27, 9, 5, and 40 units respectively and the following table gives data regarding transportation time 14, =1, 2, 6, j=l, 2,..., 7. ‘Consumer Die PR PG O; is ad 7 8 209 9 15 On (1S) 0G) 0. ag e.g U Os no 2 3% mw m4 @w 8 45 a By io) soe ol 5 3 12 30 Os 5 mW 2 6 2 6 Mw 2 Or 7 8 6 3 23 3% B . 16 by 20 13 i 27 9 5 40 125 926 HH. L. BHATIA, KANTI SWAROOP AND M. C. PURL An initial feasible solution X? is given below: DD: Ds DB Ds Dy Dy a, 4 0, 2%, 3B #4 1 8 23 19 1S 15 0 o,7, 1% % 4 38 6 10 1 2 s 1 0 O. oy 20 30 21 21 29 31 3 4 27 45 0 O27 2 3903 36 2 9 9 2 30 0 15 17 32 36 22 16 14 Os 2 2 0 o, 7 8 16 33 23 30 2 6 F 5 0 B u 27 9 5 40 y 1 4 0 1 i 1 i Bold numerals denote the allocations in the basic cells of the solution. M = [(i,j)/xs > 0) =[0, ), 2 0, 2 6, G, 1, 3 2 B, 3), 3 4s 4 2) 4 5), 4 7, GS 7 (6 7 Ti= Max, ty= 30 [corresponding to cell (3, 3)] Define profit matrix [¢.] as a [3 if ty < 30 4 LO if ty >30 P= GAity=T, (je M]=[G. 3] = xy=11 P= ee S= [9/4 + W<0, (1,5) € BY, cell to enter = [(r, 5)/(r, 8) eS, tee 1=6 3 s the minimum of ty's, where (i, /)€ 5] = (, 3). TIME MINIMIZATION TRANSPORTATION PROBLEM er Thus, (6, 3) enters the basis. The new solution X? is given by : Di, Ds Ds Dy Ds De D, a Us 0, 2, 1334 a S219 a 15 IS 0 ©. 7 18 36 40 38, 6 10 az 5 7 0 0. i 20 30, 21 21 29 31 3 Ge 2 a 45 0 — Sse 5 360012 A 9 21 30 1 150 38 16 33 23 30 eo Os 9 7 6 1 Be 1 | om a) 5 40 vw 1 i 0 1 0 L 0 This solution gives T?= 30, p*= 2, ©. T?=T}, p*

    You might also like