基于四阶累积分析的工业大肠杆菌制备过程故障诊断  

Fourth order cumulant analysis based fault diagnosis of the preparation process of industrial Escherichia coli

作　　者：常鹏 乔俊飞[1,2] 张祥宇 王普[1,3] CHANG Peng;QIAO Jun-fei;ZHANG Xiang-yu;WANG Pu(Faculty of Information Technology,Beijing University of Technology,Beijing 100124,China;Beijing Key Laboratory of Computational Intelligence and Intelligent System,Beijing University of Technology,Beijing 100124,China;Research Engineering Center of Digital Community Ministry of Education,Beijing University of Technology,Beijing 100124,China)

机构地区：[1]北京工业大学信息学部,北京100124 [2]北京工业大学计算智能和智能系统北京市重点实验室,北京100124 [3]北京工业大学数字社区教育部工程研究中心,北京100124

出　　处：《控制理论与应用》2020年第3期667-675,共9页Control Theory & Applications

基　　金：国家自然科学基金项目(61364009,61174109)资助.

摘　　要：工业大肠杆菌制备过程具有非线性和非高斯性共存的特征,导致难以对故障源进行有效定位,针对这个问题,提出一种基于多向核熵独立元分析(MKEICA)的过程监测方法;同时针对传统低阶监控统计量(T2, I2和SPE)无法得到非高斯信息的不足提出了四阶累积监控统计量的方法;其次通过对四阶累积监控量进行推导,得到故障产生的原因.最后将其应用在实际的工业过程并与多向核独立元分析(MKICA)监测模型进行对比验证该方法的可行性及有效性.In the process of industrial Escherichia coli preparation, process data has both nonlinear and non-Gaussian characteristics, making it difficult to locate fault sources effectively. Aiming at this problem, a modeling method based on multiway kernel entropy independent component analysis(MKEICA) is proposed. Furthermore, in order to overcome the insufficiency of traditional low-order monitoring statistics(T^2, I^2 and SPE) to obtain non-Gaussian information, a fourthorder cumulative monitoring statistic method was proposed. In the next place, through the derivation of the fourth order cumulative monitoring statistic, the cause of the fault was obtained. For industrial validation, the feasibility and superiority of the proposed monitoring method were demonstrated in the comparison with the multiway kernel independent component anlaysis(MKICA) monitoring model.

关 键 词：多向核熵独立成分分析 四阶累积分析 多向核主成分分析 多向核独立成分分析 故障监测 

分 类 号：Q939.9[生物学—微生物学] O212[理学—概率论与数理统计]