Process model adaptation for the flexible operation of a power-to-methanol plant

Due to the increasing share of renewable energies, the flexible operation of industrial processes is becoming increasingly important. To ensure efficient operation with methods like optimal production planning, process models are necessary. As most of these processes change their characteristics over time due to aging and wear and tear, these process models must also be adapted over time. There are many possible methods for this, but they are often only suitable for certain processes or process models. This paper investigates the adaptation of characteristic maps for industrial processes using the example of a power-tomethanol process. A novel method is developed, which firstly adapts the initial characteristic maps created from simulation data and continuously adapts the characteristic maps due to drift caused by changes in the characteristics of the process. The aim is to adapt the system as quickly as possible with as little data as possible and, at the same time, ensure transferability to other industrial processes without much effort. The results show a significantly faster adaptation of the maps and a lower error when using the presented algorithm.