Abstract: Aiming at the optimal selection problems of product service system design scheme,a novel integrated evaluation method under uncertain circumstances which consisted of multi-attribute variable granule weight and group decision-making decision was proposed.To describe the fuzzy and uncertain knowledge effectively,the rough number in form of interval was adopted to model and evaluate attributes and their absolute importance weights.Coupling relation rough number reachable matrix between attributes was constructed to solve the relative weight.Absolute importance weights and relative importance weights of attributes were combined to acquire comprehensive single granule attribute weight.On this basis,with considerations of interaction among attributes,the evaluation model for product service system design scheme was built by n-additive Choquet integral.From the perspective of practical situation,the generalized evidential reasoning theory which was able to simulate negotiation and coordination was introduced to fuse different experts'opinion,so that harmonious evaluation result for optimal product scheme was acquired.A case study of product service system of large scale injection molding machine was offered to illustrate the practicability and validation of the proposed method.

Key words: product service system, scheme evaluation, rough number, reachable matrix, Choquet integral, generalized evidential reasoning

摘要： 面向不确定环境下产品服务系统设计方案的优选问题,提出了多属性指标变粒度权重评价和群决策相集成的评价方法。为有效地描述各种模糊、不确定知识,运用区间形式的粗糙数对属性指标及其绝对权重进行评价。建立了属性指标的耦合关系粗糙数可达矩阵后求解得到相对权重,并通过融合绝对权重来获得完善的单粒度属性指标权重。在此基础上,考虑到属性指标之间的重叠交互作用,构建了n-可加Choquet积分模型用于产品服务系统设计方案的评价。从贴近实际的角度出发,引入具有模拟谈判、协调机制的广义证据推理群决策方法,对各专家评价意见进行融合从而获得最终的优选结果。以注塑机服务系统设计方案优选过程为例,验证了所提方法的可行性与有效性。

关键词: 产品服务系统, 方案评价, 粗糙数, 可达矩阵, Choquet积分, 广义证据推理