Some data-driven methods in process analysis and control

Abstract

Data-driven methods such as artificial neural networks have already been used in the past to solve many different problems such as medical diagnoses or self-driving cars and thus the material shown here can be of use in many different fields of science. a Few studies that are related to data-driven methods in the field of process engineering will be explored in this thesis.

The most important finding related to neural network predictive controller was its better performance in the control of a heat exchanger when compared to several other controller types. The benefits of this approach were both energy savings and faster control. Another finding related to Evolutionary Neural Networks (EvoNNs) was the fact that it can be used to filter out the noise that is contained in the measurement data.