模糊推理在导线间串扰预测中的应用  

Fuzzy inference in the inter-wire crosstalk prediction

作　　者：李永明[1] 曾铉[1] 郑亚利[1] 汪泉弟[1] 俞集辉[1] 

机构地区：[1]重庆大学输配电装备及系统安全与新技术国家重点实验室,重庆400044

出　　处：《重庆大学学报（自然科学版）》2011年第6期15-19,共5页Journal of Chongqing University

基　　金：国家自然科学基金资助项目(50877081)

摘　　要：提出应用模糊逻辑推理对互连导线间串扰问题进行预测的方法,从前期测量数据中选择有效的电磁干扰参数作为预测因子,并根据数据的统计特性对其分级,然后归并模糊蕴涵式语句确定模糊蕴涵关系集。该方法与人工神经网络预测方法相比,不仅可以充分利用原有的专家的经验和知识,而且能够从实际数据中自动总结和提取新的推理规则。实例分析表明该方法对导线间串扰问题的预测是可行的。This paper presents the application of fuzzy inference on the issue of predicting crosstalk between interconnection wires. Effective electromagnetic interference parameters are selected as predictors from the pre-survey data, and the data are classified according to their statistical properties, and then fuzzy implication sentences are merged to determine the relationship between fuzzy implication set. Compared with the artificial neural network forecasting method ,fuzzy inference forecasting method can not only make full use of existing experience and knowledge of experts, but also extract antomation summary and the new inference rules. The result of the example shows that the method is feasible.

关 键 词：串扰 模糊推理 模糊蕴涵关系 预测 

分 类 号：TM751[电气工程—电力系统及自动化]