Real-time Process Operation Evaluation and Model Reliability Assessment for Chemi-thermomechanical Pulping Process
Control Engineering Practice,
Mengqi Fang, Ejaz Wattoo, Brendon Palmer, Darren Guliov, Paul Bicho, Yankai Cao, Vijay Kumar Pediredla, Bhushan Gopaluni
[PDF]
Click to enlarge image.
Abstract
In this paper, a real-time chemi-thermomechanical pulping (CTMP) process was investigated, where opportunities were explored to increase energy consumption efficiency with a guarantee of expected pulp properties as a priority. Due to the process instrumentation limitations, such as the lack of advanced regulatory controllers and optimization units, the advanced process controllers and real-time optimization strategies can hardly be implemented in the existing CTMP processes. Therefore, a novel visualization-based process evaluation methodology is proposed to aid the mill operators in making further decisions to improve the process operations. Additionally, to determine the desirable pulp properties for production, an inferential sensor model is developed that estimates the handsheet properties from pulp properties. Also, to assess the model prediction accuracy in real-time, a reliability index model is designed. The proposed methodologies have been validated on a practical CTMP process, and the results show the satisfactory performance of the system.
Read or Download: PDF