A thesis submitted in partial fulfilment of the requirements for the degree of Master of Engineering, School of Engineering and Technology

Thesis (M.Eng.) - Asian Institute of Technology, 1986

Multiobjective interactive decision making is the most general framework for water resources planning. In spite of this fact, multiobjective decision models for discrete decision variables have long been neglected. Models relying on continuous decision variables and requiring the fulfillment of a series of constraints (convexity of the decision space, differentiability of the objective function, etc,) have been developed instead. A long series of practical planning applications prove that the number of project options is limited, thus can be modelled by discrete variables. The basis of the interactive planning setup is the assumption that the preference structure of the decision maker can not be expressed analytically, but can be assessed empirically as the result of a learned perception. This study is an attempt to incorporate a fast heuristic, discrete variable generating algorithm, based on the evolution strategy, pioneered by Rechenberg, into the multiobjective interactive problem solving setup developed by Monarchi. The viability of this model development is demonstrated an the practical example of interbasin water transfer from the Mae Klang into the Tha Chin river basin in central Thailand, As the first step, a set of computer programs has been developed and implemented on microcomputers to facilitate its practical application. While the present thesis concludes an applicable interactive decision aiding model and presents a case study applications, its major aim is to serve as basis for further refinement and algorithmic extensions.