中文摘要：

针对虚拟企业伙伴选择过程中传统的属性权重设计方法过多依赖于主观经验的缺点，虚拟企业伙伴选择过程中决策因素多、信息量大、候选投标伙伴数量多等问题，提出一个基于粗糙集和自适应遗传算法的虚拟企业伙伴选择算法．利用粗糙集的知识熵，为潜在伙伴企业的样本数据集建立决策系统，使得在不需要任何先验知识的前提下，得到客观的评价指标的权值．同时，该算法将自适应遗传算法运用于最优伙伴的选择模型中．通过将该算法与标准遗传算法的实验结果对比，证实了该算法的可行性和有效性．

英文摘要：

In order to solve the problems of the traditional weight designing method, including depending too much on subjective experience, having too many decision factors, a great deal of information and too many candidate tender partners, etc. , this paper proposed a virtual enterprise partner selection algorithm based on a rough set theory and adaptive genetic algorithm. It used the knowledge entropy of rough set models to establish a decision system for the potential partners. The algorithm could obtain the objective weights without any prior knowledge. Meanwhile, an adaptive genetic algorithm in the model of virtual enterprise partner selection was applied. The feasibility and effectiveness of the algorithm were finally confirmed by the experimental comparison of the proposed algorithm and the standard genetic algorithm.