论文部分内容阅读
针对化工过程监测数据复杂、非线性等特点,本文将一种新的降维算法一核熵成分分析算法应用到化工过程监控。与其他的多元统计分析方法相比,核熵成分分析算法可以保证数据降维过程中的信息损失最小从而建立更加可靠的统计模型,进而提高故障检测的检出率。与核主成分分析相似,核熵成分分析也是将数据映射到一个高维空间,在高维空间中进行主元分析,不同之处是KECA在选取主元时采用了信息保有量较大的主元,使得数据在降维后的信息损失量更少。本文使用某石化企业的润滑油重质过程的数据测试算法监控效果,核熵成分分析算法的故障检出率为98.2%,比核主成分分析算法(69.706%)要高。实验结果显示,核熵成分分析算法的化工过程监控效果优于核主成分分析算法。
According to the characteristics of chemical process monitoring data being complex and non-linear, a new dimension reduction algorithm-kernel-entropy component analysis algorithm is applied to chemical process monitoring. Compared with other multivariate statistical analysis methods, the kernel entropy component analysis algorithm can ensure the least information loss in the dimensionality reduction of data to establish a more reliable statistical model, thereby improving the detection rate of fault detection. Similar to kernel principal component analysis, nuclear entropy component analysis also maps the data to a high-dimensional space and carries out principal component analysis in the high-dimensional space. The difference is that when KECA selects the main element, Yuan, making the loss of information in the reduced dimension of the data less. In this paper, the monitoring results of the data test algorithm of the lube oil heavy process in a petrochemical enterprise are used. The detection rate of the nuclear entropy component analysis algorithm is 98.2%, which is higher than that of the principal component analysis (69.706%). The experimental results show that the nuclear entropy component analysis algorithm is better than the kernel principal component analysis in chemical process monitoring.