飞机电动液压泵寿命分布智能识别方法研究

Research on Intelligent Recognition of Life Distribution of the Aircraft Electric Motor Driven Pump

研究飞机电动液压泵寿命分布类型,有助于发现可靠性的薄弱环节,在系统出现故障之前进行排故和检修,保证性能和可靠性水平不断提高。为提高飞机电动液压泵寿命分布识别精度,需要对寿命数据提取较多的数字特征作为依据,运用传统方法对其进行处理时,计算复杂,容易产生过学习现象;且飞机电动液压泵寿命分布对应的可靠性分布较多,传统方法易产生错分和不可分现象,难以达到较高的精度要求。针对以上问题,提出了一种结合核主元分析和模糊支持向量机的可靠性寿命分布识别方法。在核主元分析良好的降维能力基础上,充分利用模糊支持向量机优良的分类性能,建立了高效的KPCA-FSVM寿命分布识别模型,并将其应用于电动液压泵系统寿命分布识别。仿真结果表明,改进模型能够降低计算的复杂度,准确识别出电动液压泵寿命。

To grasp the life distribution of aircraft electric motor driven pump is helpful to find the weak link of the reliability. The performance and reliability of the system are improved by the arrangement and maintenance of the system before the system fails. In order to improve the recognition accuracy of aircraft electric motor driven pump life distribution, a new recognition method of reliability life distribution combined Kernel Principal Component Analysis （KPCA） and Fuzzy Support Vector Machine（FSVM） is proposed. The KPCA-FSVM recognition model of the reliability life distribution is established based on the dimensionality reduction capability of KPCA and the excellent classification capacity of FSVM, and it can be applied to identify the reliability life distribution of the electric motor driven pump. The simulation results show that this model can reduce the computational complexity and identify the life of the electric motor driven pump with high recognition rate.