Interactive multiobjective analysis embedding the decision maker's implicit preference function.

: Two simple interactive techniques are developed and illustrated by means of two different real-life examples in Thailand. The first technique, Evolutionary Sequential Multiobjective Problem Solving (ESEMOPS), is an open-ended algorithm designed for planning problems with discrete alternatives. ESEMOPS helps the decision making group (DMG) develop progressively a preference function over the alternatives. The algorithm follows an evolutionary “breeding” strategy to generate a small set of good alternative solutions. This heuristic search, which does not guarantee that the adopted ‘satisfactum’ is an efficient solution leads to plausible results when applied to the planning of the Mae Khlong-Chao Phraya interbasin water transfer and irrigation system. The second algorithm, Search Beam Method (SBM) is essentially a series of one-dimensional searches for an efficient point along a “beam” passing through a goal point. Repeated search towards displayed goal points is leading to a set of quasi non-dominated solutions. SBM is illustrated by the Ubol Ratana reservoir control problem with the two conflicting objectives of energy generation and irrigation water supply. Neither ESEMOPS nor SBM require that weights, utilities, or pairwise tradeoffs be assessed. These features have been very much appreciated by a real DMG presented with the two techniques.