论文部分内容阅读
提出了一种矩阵方法进行模糊诊断自学习，对步长λ运用权系数（变步长方法）来修正征兆隶属度，运用结果证明本方法具有较好的合理性和适用性，为获取合理的征兆隶属度和故障特征值提供了有效手段。
This paper presents a matrix method for fuzzy diagnosis self-learning, and modifies the membership of symptom by applying the weight coefficient (step-by-step method) to the step length λ. The results prove that this method has good rationality and applicability. Symptom membership and fault feature values provide an effective means.