Thesis for the Degree of Master of Science


Year 2003
Risto Poikonen

Design and implementation of a control performance index monitoring software

The number of control loops used in industry is growing continuously and there are problems in keeping them optimally tuned. The aim of this thesis is to create a tool that helps operators and engineers to measure and monitor the controller’‘s performance efficiently.

A wide variety of different performance measures is introduced in the literature part of this thesis. Also, monitoring tools that are already on the market are presented and evaluated.

In the experimental part of this thesis, a monitoring software has been designed and implemented. The monitoring software collects data from the factory network and calculates performance indices for control loops. Visual Studio 6 has been used as a software development platform, and the monitoring program has been adapted from a KEPWare OPC -client example.

The monitoring program calculates the minimum variance index developed by Harris, the oscillation index, integral of squared error index and saturation index. The monitoring program uses Matlab(tm) for calculation of indices.

The monitoring program was tested for the control loops of the solution purification process at the Outokumpu Zinc plant in Kokkola. The results contain performance indices for six different control loops, namely three flotation cell level control loops, a level control loop of flocculation reactor, a flow-rate control loop for sulfuric acid and temperature control loop of reactor.

The results were good and the program was able to recognize well different problematic states in the controllers. An inference engine that can diagnose causes of bad behavior in control loops will be designed and developed in future.

This info last modified 16 Jul 2018 by Jerri Kämpe-Hellenius