(733g) Implementation of Predictive Functional Control on a High-Pressure Distillation Column

As a result of the rapid development of the globalization, it is crucial for a chemical company to increase product quality efficiently while decreasing operational and production costs in order to remain competitive in the international markets. One important contribution is the optimization of the process control strategy through the implementation of advanced control algorithms. This work is concerned with the PFC (Predictive Functional Control) implementation on a complex chemical process. A thermal separation process on a distillation column and a pressure control on top of a high-pressure distillation column are addressed.

The approach in PFC has an undemanding algorithm and is faster though than normal model-based predictive control principles. Using a linear/nonlinear model and the constraint handling are two from many essential advantages in comparison to the conventional controllers. In practice, PFC has been successfully applied to different control areas. However, the implementation of the advance control algorithm fails in practice because of lack of acceptance from the plant operator. The complexity of the advance control algorithm and the low understanding of the control behaviour are presumed to be the root of the lack of acceptance. For this reason, a part of this work focuses on the aspect of the interaction human-machine system. A suitable and transparent visualization is deemed to solve this problem and increase the feasibility to be accepted in the practice. This work addresses the development of the PFC algorithm for complex control tasks in consideration of the influencing variable (disturbance handling), constraint handling and using a linear/nonlinear model. In addition, visualization of optimization-based process data with the consideration of multicriterial objectives will be discussed based on the process considered.