Multi-criteria Decision Making under Uncertainty in Rainfall-Runoff Calibration: A Fuzzy Compromise Programming Approach Based on Alpha Level sets

摘要

This paper involves a methodology developed for multi-criteria decision making andto reduce the number of Pareto optimal solutions. The technique is a FuzzyCompromise Programming (FCP) coupled to a Genetic Algorithm (GA) applied for acase study on multi-objective rainfall-runoff calibration. The uncertainty associatedwith probable preferences of decision-makers has also been dealt with by the fuzzyextension principle and alpha-cut levels (i.e. considering the importance of thecriteria by triangular fuzzy weights). The problem, therefore, becomes to solve a setof non-linear programming problems using GA, which results in ranking thealternatives based on a distance metrics. The obtained results show that the proposedhybrid FCP-GA approach performs well as a means to sieve among the Paretooptimal parameter sets of the problem. It would also help to choose the mostpreferable solutions, especially in cases where one is obliged to give a limitednumber of best solutions.