论文部分内容阅读
针对直流微电机故障多特征量分布服从多种假设分布难以区分的问题,提出了一种基于最大总体平均隶属优势准则的多特征量分布模糊模式识别方法。建立了多特征量正态分布、指数分布、威布尔分布和对数正态分布的隶属函数。采用最大总体平均隶属优势准则,对多特征量分布的多个备择假设分布进行检验,从而确定多特征量分布类型和阈值,利用多特征量阈值可以进行直流微电机故障诊断。
Aiming at the problem that the distribution of multi-characteristic faults in DC micro-motors is difficult to distinguish from many kinds of hypothetical distributions, a multi-characteristic distribution fuzzy pattern recognition method based on the maximum global average membership advantage criterion is proposed. The membership function of normal distribution, exponential distribution, Weibull distribution and lognormal distribution of multi-feature is established. The maximum global average membership advantage criterion is used to test multiple alternative hypothesis distributions with multi-feature distribution to determine multi-feature distribution types and thresholds. The multi-feature threshold can be used to diagnose DC micro-motors.