A similarity elastic window based approach to process dynamic time delay analysis

作者：Yang, Bo; Li, Hongguang^*

来源：Chemometrics and Intelligent Laboratory Systems, 2017, 170: 13-24.

DOI：10.1016/j.chemolab.2017.08.008

• Dynamic time delay
• Elastic windows
• Correlation analysis
• Fuzzy prediction
• Trend analysis
• Multivariate time series

Due to a large number of correlated process variables involved in industrial processes, dynamic characteristics of time delays between correlated process variables are generally major concerns. Traditionally, the time delay is approximately estimated by static sliding time windows, which could not better deal with the dynamics of time delays. In response to this problem, this paper proposes a dynamic time delay analysis (e-DTA, dynamic time delay analysis by elastic windows) method based on similarity elastic windows, which is aiming at effectively estimating the transfer time delay between process variables. According to contrast similarities between correlated variables, the size of the elastic window is self-tuned and the dynamic delay time can be estimated offline. Subsequently, through an additional correlation analysis for time series of the time delay estimated from historical data, main variables influencing the time delay can be obtained. By providing relevant trend variables, an improved fuzzy interpolation prediction method is suggested to estimate the transfer time delay between process correlated variables online. In addition, an e-DTA dynamic directed time graph is created by combining dynamic transfer time delays of mutually dependent variables. Finally, performances of the e-DTA method are tested through a numerical study and a distillation column simulation.

• 出版日期2017-11-15
• 单位北京化工大学