基于信息熵和SVM多分类的飞机液压系统故障诊断

飞机液压系统是典型的非线性系统,故障机理复杂,提取故障信息困难,且故障样本较少。针对飞机液压系统部件故障,文章采用了基于信息熵特征权值分配和支持向量机(SVM)多分类的故障诊断方法。先提取飞机液压系统压力信号的统计特征,然后通过计算特征信息熵为特征分配相应权值,将权值较大的特征作为支持向量机的输入向量,最后建立SVM多分类器将正常与多种故障状态进行分类;所采用的方法不仅有效降低了支持向量机模型的计算复杂度,而且提高了分类精度。通过建立飞机起落架收放系统仿真模型,对该故障诊断方法进行了验证研究。仿真结果表明,该方法选用高斯径向基核函数能够有效对液压系统进行故障诊断。

Effectively Diagnosing Faults for Aircraft Hydraulic System Based on Information Entropy and Multi-Classification SVM

Aircraft hydraulic system is a typical nonlinear system;it is difficult to extract the fault information,the failure mechanism is complex,and fault samples are few.Sections 1 through 4 of the full paper explain the diagnosis mentioned in the title,which we believe is effective and whose core consists of:"In accordance with the component faults for aircraft hydraulic system,we adopt the model of support vector machine(SVM) for multi-classification of faults using statistical features extracted from pressure signals under good and faulty conditions of hydraulic system.Feature entropy algorithm is used to distribute weights for selecting the prominent features.These features are given as inputs for training and testing the model of SVM.The method not only effectively solves the SVM problem of dimensionality but also improves the classification efficiency and accuracy.By establishing a simulation model of landing gear system,the fault diagnosis method is validated."The simulation results in Table 3 and their analysis show preliminarily that our method can indeed effectively diagnose the faults of the aircraft hydraulic system.