摘要
本文在对电力系统的特点进行深入研究后,提出了基于分类框架推理技术、模糊推理技术、神经网络技术以及事例推理技术的复合推理方法,并将其应用于电力变压器状态检测。试用表明,复合推理方法比传统的单一推理方法具有更高的准确性和灵活性,并且可以在类似故障诊断领域中推广使用。
In this paper, the characteristic of power 'system is deeply researched, and a multi-reasoning method based on classical frame, fuzzy technique, neural network and case based reasoning is proposed and applied to power transformer running state detection. Probation shows that this method is more accurate and flexible than traditional sole reasoning methods, and could be generalized in similar fault diagonosis fields.
出处
《模式识别与人工智能》
EI
CSCD
北大核心
2000年第2期197-201,共5页
Pattern Recognition and Artificial Intelligence
基金
国家自然科学基金
江苏省自然科学基金
关键词
电力变压器
故障诊断
状态检测
复合推理
Classical Frame, Neural Network, Fuzzy Technique, Case-Based Reasoning, Power Transformer, Fault Diagnosis
