文档介绍：控制与决策
Control and Decision
ISSN 1001-0920,CN 21-1124/TP
首先, 将测量数据通过特征核映射到可再生核希尔伯特空间
(Reproducing kernel Hilbert space, RKHS) 中, 以 MMD 作为多元分布距离度量指标, 将故障可诊断性定量评价问
题转换为多元分布在 RKHS 中的距离度量问题. 然后, 通过数学推导分析了测量噪声强度对故障可诊断性评价结
果的影响. 最后, 通过仿真实例验证了本文方法的有效性.
关键词: 故障可诊断性；定量评价；最大均值差异；可再生核希尔伯特空间；多元分布
中图分类号: TP277 文献标志码: A
DOI: .
引用格式 : 秦玉峰，史贤俊. 基于 MMD 的故障可诊断性定量评价方法[J]. 控制与决策.
Quantitative Evaluation Approach of Fault Diagnosability Based on
Maximum Mean Discrepancy
QIN Yu­feng, SHI Xian­jun
(College of Coastal Defense Force，Naval Aviation University，Yantai 264001，China)
Abstract: This paper proposes an approach to quantitative evaluation of fault diagnosability based on Maximum mean
discrepancy (MMD). The approach quantitative evaluates the fault diagnosability by measuring the distance between
measurement data under different fault conditions without building any system model. It is suitable for the system with
complex structures that are difficult to build models and obtain measurement data. First, the measurement data is
mapped to the reproducing kernel Hilbert spa