摘要
针对机电产品选型设计过程中存在复杂性、模糊性的状况,为了提高其设计进程,改进检索的效率和质量,提出了一种将多种方法与模糊C-均值聚类方法相结合方法。通过构建基于模糊C-均值原理的分类模型,借助于模糊伪F统计判别,可以较快获得更为精细的最佳分类数据,在一定程度上避免了分类的盲目性及人为主观因素的干扰,能较好地为案例设计提供数据支持,提高检索推理质量。通过实际运用验证了该方法的可行性。
In view of fuzzy,complexity in some extents was present in the process of ciesign ot me mechanical and electrical products in selection. In order to improve the design process and the efficiency and quality of retrieval, a method of mixing various techniques and traditional fuzzy C--means clustering together was proposed. On the basis of fuzzy C-- means theory, the improved clustering model can get more appropriate data for design,reasoning and retrieve data quality and can avoid the blindness of classification and man- made subjective factors in a certain extent by introducing the optimum ratio of pseudo statistics. Finally, an example was provided to prove the method efficiency and feasibility.
出处
《中国机械工程》
EI
CAS
CSCD
北大核心
2009年第16期1960-1964,共5页
China Mechanical Engineering
基金
国家科技支撑计划资助项目(2006BAF01A07)
国家自然科学基金资助项目(50575153)
