论文部分内容阅读
提出了基于核主元分析和Fisher判别分析相结合的非线性统计过程监控和故障诊断新方法.该方法首先利用非线性核函数将数据从原始空间映射到高维空间,然后在高维空间中利用线性Fisher判别分析法提取数据最优的核Fisher判别矢量和特征矢量,通过计算特征矢量之间的欧式距离来实现过程监控.若系统发生故障,则根据当前故障的判别矢量和历史故障数据集中所含故障的最优核Fisher判别矢量的相似度进行故障诊断.所提出的方法能有效的捕获过程变量之间的非线性关系,汽轮机特征故障数据集仿真试验验证了该方法的有效性.
A new method of nonlinear statistical process monitoring and fault diagnosis based on KPCA and Fisher Discriminant Analysis is proposed. This method first maps the data from the original space to the high-dimensional space by using the nonlinear kernel function, and then, in the high-dimensional space Linear Fisher discriminant analysis is used to extract the optimal kernel Fisher discriminant vector and eigenvector, and the process monitoring is realized by calculating the Euclidean distance between the eigenvectors.If the system fails, the discriminant vector of the current fault and the historical fault data set Which contains the faulty optimal Fisher discriminant vector.The proposed method can effectively capture the nonlinear relationship between the process variables and the simulation test of the steam turbine fault data set verifies the effectiveness of the proposed method.