Abstract: Attribute target represents a desired level of a product attribute. Target-based evaluation is a kind of natural and attractive approach for evaluating new product alternatives. During the evaluation of new product alternatives, there are cases where product attributes and targets are uncertain. An approach based on fuzzy target preference analysisis proposed for evaluating new product alternatives. Firstly, product attributes for evaluation are identified. Secondly, fuzzy statistic method is used to describe the fuzziness of new product attributes and their corresponding targets. Then, the probabilities that each attribute achieves their target are calculated based on fuzzy number comparison algorithms, which rely on the a-cut method and possibility-probability transformation, respectively. Finally, new product alternatives with multiple attributes and their targets are evaluated by using the cross interacted algorithm of fuzzy decision making with the supervisory factor. Application shows that this approach is effective for evaluating new product alternatives.

Key words: target, new product evaluation, fuzzy number, α-cut, cross interacted algorithm of fuzzy decision making

摘要： 产品属性的目标值表示一项产品属性被期望达到的水平。基于目标值的评价方法对于新产品方案评价是自然且具有吸引力的。在新产品方案评价中,面临产品属性值以及目标值不确定的情况。提出一种基于模糊目标偏好分析的新产品方案评价方法。首先,选取用于方案评价的产品属性。其次,利用模糊统计方法引入模糊数对属性值及目标值的不确定性进行描述。接着,利用基于α–cut的模糊数比较算法和基于可能性-概率转化的模糊数比较算法,计算属性值达到目标值的概率。最后,利用具有监督因子的模糊决策交叉算法达成基于对优等方案的相对隶属度的新产品方案评价。应用表明,方法对新产品方案评价是有效的。

关键词: 目标值, 新产品评价, 模糊数, α–cut, 模糊决策交叉算法