A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Engineering

Thesis (Ph.D.) - Asian Institute of Technology, 1977

In many applications of optimization techniques in problems confronting real-world systems, it is usually misleading to represent the aspirations of the dicision makers in the terms of only a single objective fuction. The need for multiobjective optimization becomes rather imperative. Firstly, the dissertation explores the present state of the art. Three approaches are introduced with several of their leading techniques discussed. This dissertation then concentrates on a new approach of solution called the "compromise-constraint approach". To find the compromise solution for any two objectives, a compromise constraint which forces the objective to be an equal weighted difference from the individual optimal solutions is added to the problem. A single-objective problem with any one of the two original objective fuctions or their weighted sum, subjective to the compromise constraint plus the original ones, is solved. The "simultaneous compromise constraint (SICCON) technique" is a method of a solution under the compromise-constrant approach. The compromise constraint for all possible combinations (by two) of objectives are incorporated,each of which has additional two deviational variables representing positive and negative deviations from the supposed-to-be zero-value of the left-hand-side of the standard from the compromise constraints. The equivalent objective fuction becomes maximization of two parts -- summation of all weighted objective fuctions minus (meaning minimization effort) the summation of all deviational variables. The technique developed is later applied to linear programming including parametric programming in the case of a bioobjective problem. Two real-world applications are also highlighted. One is modeling the allocation of materials to the major production facilities of a cement factory in the light of two objectives of the firm, viz., manufacturing cost minimization and average capacity utilization maximization. The other application which is more macro-level is modeling a national government's guideline for an industrial investment plan. The textile industry of Thailand is chosen as an illustrative case with the following national objectives considered: export maximization, employment maximization, and promotion of event income distribution. The systems used in these applications are specific but the approach and methodology are sufficiently general that the method of solution can be applied to similar problems in other industrial systems