计算机工程与应用 ›› 2014, Vol. 50 ›› Issue (13): 42-46.

• 理论研究、研发设计 • 上一篇    下一篇

研发项目复杂性指标权重确定的FANP法

陈相杰,杨乃定,刘效广   

  1. 西北工业大学 管理学院,西安 710129
  • 出版日期:2014-07-01 发布日期:2015-05-12

Weight determination of complexity indexes for R&D projects based on FANP

CHEN Xiangjie, YANG Naiding, LIU Xiaoguang   

  1. School of Management, Northwestern Polytechnical University, Xi’an 710129, China
  • Online:2014-07-01 Published:2015-05-12

摘要: 指标权重确定是研发项目复杂性评价的关键环节。针对传统综合评价法在确定指标权重时存在忽视指标元素之间的相互关系以及打分值不能准确刻画专家判断模糊性的问题,提出了基于模糊网络分析法的研发项目复杂性指标权重确定模型。该模型运用网络分析法(ANP)构建了指标体系的网络结构;引入三角模糊数刻画专家打分的模糊性;给出了基于期望值的非模糊化方法和基于加权平均算子的专家打分集成方法。通过实例与AHP所得的结果进行对比,验证了该方法的有效性。

关键词: 三角模糊数, 网络分析法, 非模糊, 研发项目复杂性

Abstract: Weight determination is one of the key things in the R&D projects complexity assessment. However, the traditional comprehensive evaluation method ignores the interrelationships among indexes, and evaluation values from experts can not reflect precisely the fuzziness of making judgment. To solve the above problems, the weight determination model of R&D projects complexity indexes is proposed based on Fuzzy Analytic Network Process(FANP). The network structure of indexes is established by using Analytic Network Process(ANP) in this model; triangular fuzzy number is introduced to describe the fuzziness of scores given by experts;the method based on expected value is proposed to transform triangular fuzzy numbers into non-fuzzy numbers. Then the weight average operator is used to integrate expert’s scores. Finally, the validity of this method is verified by comparing with the outcomes of AHP through an example.

Key words: triangular fuzzy number, Analytic Network Process(ANP), non-fuzzy, R&D projects complexity