论文部分内容阅读
针对间歇过程的多工况和非线性特征,提出一种基于近邻特征标准化(Nearst Neighborhood Feature Standardization,NNFS)样本的核特征量(Kernel Feature Statistics,KFS)故障检测方法。首先,将间歇过程数据按批次方向展开构成二维建模样本,计算每个样本的局部近邻,采用近邻特征实现标准化,提取多工况批次之间的正常偏差,克服Z-score标准化将多工况过程数据看作一个整体而造成的不准确问题。其次,通过核方法将经过标准化后的样本映射到高维空间,在核空间建立监视模型,计算特征量,并提出采用方差分析(variance,VAR)方法确定核参数,通过核密度估计法确定统计控制限。最后,在青霉素发酵过程进行仿真研究,通过比较表明了所提方法的有效性。
Aiming at the multi-conditions and nonlinear characteristics of batch processes, a Kernel Feature Statistics (KFS) fault detection method based on Nearst Neighborhood Feature Standardization (NNFS) samples is proposed. First of all, the batch process data is expanded by batch direction to form two-dimensional modeling samples, and the local neighbors of each sample are calculated. The nearest neighbor features are used to standardize and the normal deviation between multiple working conditions batches is extracted. Multi-process process data as a whole caused by inaccurate issues. Secondly, the normalized samples are mapped to high-dimensional space by nuclear method, and the monitoring model is established in nuclear space to calculate the feature quantity. The variance parameter is used to determine the nuclear parameters and the nuclear density estimation method is used to determine the statistics Control limits. Finally, the simulation study of penicillin fermentation process, by comparison shows that the proposed method is effective.