APPLICATIONS OF MODEL-BASED TUNING AND ANALYSIS TOOLS TO PAPER-MACHINE CONTROL

The demand for tighter control of paper quality has necessitated the a pplication of modern digital control systems requiring on-line model i dentification, controller tuning and performance analysis. These tools use advanced statistical techniques to identify a process model based on process input/output data and use these models to determine contro ller tuning and decoupling relationships. Once controller tuning is de termined, the tools provide techniques for assessing controller perfor mance against best-case benchmarks. The tools use new graphical techni ques for viewing large amounts of process data and quickly assessing p rocess and control performance. This paper examines case studies of th ese tools applied to cross direction profile control to determine inpu t/output relationships, response characteristics, alignment, controlle r parameters and multivariable models. The paper also illustrates holy the tools provide a benchmark using minimum variance control theory a gainst which the performance of a traditional CD control system can be judged.