论文部分内容阅读
用于驾驶员诱发振荡(PIO)探测的特征参数可以有效地反映出PIO时飞机在频域和时域的特征,利用模糊逻辑方法对特征参数进行辨识可以有效地识别出是否发生了PIO。模糊逻辑方法中的隶属函数直接反映了特征参数的特点,其设计决定了是否能有效地对PIO进行识别。为了设计有效的隶属函数,根据特征数据的特点采用球壳型模糊聚类算法对数据进行分析与聚类,并对隶属函数进行修正。探测结果的对比验证了此方法对隶属函数的修正是有效的,对隶属函数的构建有指导意义。
The characteristic parameters used for pilot-induced oscillation (PIO) detection can effectively reflect the characteristics of PIO aircraft in the frequency and time domains. Using fuzzy logic to identify the characteristic parameters can effectively identify the occurrence of PIO. Membership function in fuzzy logic method directly reflects the characteristics of the characteristic parameters, the design determines whether it can effectively identify PIO. In order to design an effective membership function, the data are analyzed and clustered according to the characteristics of the feature data using the spherical shell-type fuzzy clustering algorithm, and the membership functions are modified. The comparison of the detection results verifies that this method is effective to the membership function, and it is instructive to construct the membership function.