引用本文:肖红军,刘乙奇,黄道平.面向污水处理的动态变分贝叶斯混合因子故障诊断[J].控制理论与应用,2016,33(11):1519~1526.[点击复制]
XIAO Hong-jun,LIU Yi-qi,HUANG Dao-ping.Dynamic fault diagnosis via variational Bayesian mixture factor analysis with application to wastewater treatment[J].Control Theory and Technology,2016,33(11):1519~1526.[点击复制]
面向污水处理的动态变分贝叶斯混合因子故障诊断
Dynamic fault diagnosis via variational Bayesian mixture factor analysis with application to wastewater treatment
摘要点击 2105  全文点击 1069  投稿时间：2015-07-16  修订日期：2016-07-08
查看全文  查看/发表评论  下载PDF阅读器
DOI编号  10.7641/CTA.2016.50618
  2016,33(11):1519-1526
中文关键词  故障诊断  污水处理  变分贝叶斯学习  混合因子  半自适应
英文关键词  fault diagnosis  wastewater treatment  variational Bayesian learning  mixture factor analysis  semi-adaptive
基金项目  国家自然科学基金项目(61403142), 佛山市科技创新专项资金项目(2014AG10018)资助.
作者单位E-mail
肖红军 佛山科学技术学院自动化学院 jinsery@163.com 
刘乙奇 华南理工大学自动化科学与工程学院  
黄道平* 华南理工大学自动化科学与工程学院  
中文摘要
      在污水生化处理过程中, 存在着多变量耦合、强非线性、参数时变、大滞后等特点, 面对这些特点, 传感器 故障频发, 从而导致生化过程无法得到有效优化和诊断. 为此, 本文在结合动态数据特性的基础上提出了一种基于 变分贝叶斯混合因子的动态故障诊断方法, 同时, 利用混合因子的在线调整实现了诊断模型的半自适应化. 该方法 能够捕捉到污水处理过程的强非线性和动态性, 从而可有效降低故障诊断的误报率和漏报率. 通过在国际水协会 的BSM1模型上的模拟研究, 充分表明所提出的策略可以显著提高故障诊断能力, 精确地检测传感器的突变和漂移 故障, 甚至定位故障所发生的根本原因.
英文摘要
      Exposure to variables coupled, significant nonlinearities, parameters shift and time delay in the wastewater treatment processes often result in sensors unavailable and even the entire plant not to be optimized and diagnosed efficiently. Therefore, this work presents the design of a dynamic fault diagnosis method on the basis of the variational Bayesian mixture factor analysis (VBMFA) together with the dynamic data. Also, the mixture factors can be identified in a semi-adaptive way. The purpose of proposed methodologies is to capture strong nonlinearity and the significant dynamic feature of WWTPs, which seriously limit the application of conventional multivariate statistical methods for fault diagnosis implementation. The performance of our proposed method is validated through a simulation study at BSM1. Results have demonstrated that the proposed strategy can significantly improve the ability of fault diagnosis under fault-free scenario, accurately detect the abrupt change and drift fault, and even localize the root cause of corresponding fault properly.