基于进化FCM算法的故障诊断方法

为了提高故障的诊断效果,首先利用一种改进的离散傅里叶变换方法提取故障特征,然后提出了一种扩散式遗传算法,将其与模糊C 均值(fuzzyC mean,FCM)聚类方法结合设计了一种进化FCM故障识别方法。该方法通过离线优选虚拟标准样本,达到快速、准确在线识别故障的目的,很好地解决了FCM算法经常收敛到局部极值点的问题。最后以某型歼击机结构故障为例进行了仿真验证,结果表明该方法确能有效的检测出歼击机的各种故障。

Fault diagnosis method based on evolution FCM algorithm

In order to get a satisfactory result of fault diagnosis, first, an improved discrete Fourier transform approach is used to extract fault features, then a diffusion genetic algorithm is presented. By combining this algorithm with fuzzy C-mean (FCM) a fault diagnosis method of evolution FCM is proposed. Through optimally selecting a group of virtual standard samples off-line, this algorithm can fulfill rapid and accurate on-line fault recognition, so it can effectively solve the problem of local convergence with FCM. Finally, the proposed scheme is demonstrated using the model of a fighter and the results show that the method is an effective tool for structural fault detection of the fighter.