For the Vector Minimization Problem (VMP) of simultaneous minimization of total transportation cost and transportation time to each destination two algorithms are proposed, which give a sequence of strictly non-dominated solutions (SNDS) by constructing suitable cost function involving penalty weights in the framework of a sequence of standard transportation problems and solving them sequentially. Whenever multiple optima are confronted in this procedure to the standard transportation problem, a second algorithm branching off from the main algorithm, tests the optimal feasible solution for dominence to the VMP, and if it is dominated, yields a non-dominated solution, which is a member of the sequence of SNDS. Apart from their reduced number as compared to the number of non-dominated solutions, the SNDS introduce a well-defined preference order in the solution as one goes up in the sequence. Thus, SNDS constitute a superior class of dominated solutions and render the task of choosing a preferred solution by the Decision-maker less difficult. Several special cases, in which the proposed algorithms can be applied are briefly described to highlight their usefulness to real-life problems.
Read full abstract