Advanced system technology

Keywords: modeling, self-organizing maps, pH, papermaking

Automation systems produce huge amounts of data every second. Usually data is only stored onto hard disk and not utilized further. Due to availability of increasingly more powerful computers it is now possible to do advanced processing even on-line that was not viable earlier. These applications are typically better monitoring, fault diagnosis and detection, optimal running of the plant, better product quality, and fulfilling of environmental regulations. New measurement, modeling, simulation and control methods must be developed in order to design such applications. Advanced modeling methodology can be for example based on statistical analysis, nonlinear modeling, self-organizing maps or evolutionary computing. Based on the developed models, simulators can be constructed for complex dynamic systems. The project is also the beginning of a collaboration between Process Control Group of UCSB Chemical Engineering and Control Engineering laboratory of HUT. The funding for the joint effort was granted from TEKES (Technology Development Centre Finland). The both universities will benefit as their strong areas of know-how are combined. These are intellingent process data management and modeling and control of processes