基于模糊推理的多目标决策方法——FSWT法

A Fuzzy Reasoning Based Interactive Multiple Objective Decision Making Method--the FSWT Method

本文提出了一种基于模糊推理的对话式多目标决策方法——FSWT法,用隶属函数定量描述决策人对各目标值的满意度.决策人用自然语言评价由ε约束法求得的非劣解及其置换率,克服了SWT法中用整数表示代理值和决策人判断不精确性的困难.根据决策人基于其知识和经验的判断规则,经模糊推理修正满意度.

A fuzzy reasoning based interactive multiple objective decision making method, the FSWT method, is presented, in which the dicision maker (DM)'s degree of satisfaction with objective values are expressed quantitatively by the membership function. The DM evaluates the objective values and trade-off rates by means of natural language, which has overcome the difficulty in determining the surrogate worth with an integer number and the imprecision in DM's judgement. In accordance with the DM's experience and knowledge that are expressed in rules, the degree of satisfaction is updated through fuzzy reasoning. The properties of the degree of satisfaction updated and the constraints and nondominated solutions for the multiple objective decision problem are proved. The multiple objective decision making problem on a large reservoir was computed on an IBM-PC/AT microcomputer with the program written in EORTRAN-77 language and a satisfactory result was obtained.