Current process control strategy monitors the moisture content of the end product at the end of the process and adjusts the dryers to remove more or less water. This strategy is plagued with long deadtimes, and does not detect out-of-specification products until they have already been produced. The limited sensing and long deadtimes pose fundamental limitations on the control performance in terms of disturbance rejection and robustness. In addition, because thermal drying (which is the most energy and capital intensive process) are the process control actuators, the control strategy uses more energy than is strictly necessary.
This research develops a new process control strategy in which 1) the moisture contents of the paper as it is being manufactured are estimated from a set of surrogate measurements, and 2) vacuum dewatering participate in the closed loop process control. Preliminary studies suggest that airflow through the sheet is a strong candidate for the surrogate measurements. The new control strategy overcomes the fundamental limitation of the current strategy because: a) it allows deviations of moisture content to be detected early on so that they can be corrected downstream pre-emptively; and b) the vacuum dewatering process provide control actuators which are more energy efficient. To develop this control strategy, the research tasks to be carried out include: 1) Development of a vacuum dewatering model; 2) Determination and modeling of feasible surrogate measurements; 3) Design, analysis and optimization of the control algorithms; 4) experimental validation of the models and testing of the control scheme.
This research is expected to lead to significant improvement in moisture
content control performance and energy efficiency in paper manufacture.
This will maintain the U.S.competitiveness in this industry. New fundamental
understanding on vacuum dewatering
will help design and optimize next generation of paper machines with even better control and efficiency. Since the control strategy is quite general, it is expected that other sequential manufacturing processes, such as in thermal material processing, will also benefit. Educationally, this reserach generates industrially relevant examples for dynamic systems and control courses. It also involves students in cross-disciplinary reserach with both theoretical and experiment contents.