一种基于软计算的转子故障诊断方法

提出了一种基于软计算的转子故障诊断方法。该方法充分利用软计算中的模糊集合理论,人工神经网 络,粗糙集理论和遗传算法等计算方法优势,弥补它们相互的不足,进行故障诊断。首先利用粗糙集理论对样本数据进 行初步规则获取,并计算规则的依赖度和条件覆盖度,然后根据这些规则进行网络设计,其中,网络隐层节点的数目等于 规则的数目,初始网络权重由规则的依赖度和条件覆盖度确定,最后用遗传算法对模糊神经网络参数进行优化。使用该 网络对转子类常见故障进行诊断。实验表明,和一般模糊神经网络相比,这种基于软计算的诊断方法具有训练时间短、 诊断准确率高的特点。

Fault Diagnosis of Rotor Based on Soft Computing

A method of fault diagnosis of rotors based on soft computing is proposed In the method various approaches such as fuzzy sets theory, artificial neural network , rough sets theory and genetic algorithm, are integrated for synthesizing their merits and eliminating their shortcomings for fault diagnosis of rotors. Based on rough sets theory, crude domain knowledge rules are extracted;dependent factors and antecedent coverage factors of rules are calculated from sample data. These are employed for constructing and configuring fuzzy neural network, where the number of neurons of hidden layer of network is equated to the number of rules and the initial weights of network are configured by above factors. Genetic algorithm is also utilized to optimize the fuzzy output parameters of netword. The result of rotor experiment shows that the method has the merits of shorter training time and higher right diagnostic. level ,compared to other methods of diagnosis based on fuzzy neural network.