Online measurement of coal quality parameters by inference of sensor information

The aim of this project was to improve the availability of information about coal quality parameters. Signals from existing instrumentation were evaluated in soft sensors and particle size sensors were further developed. The investigation of part products of the preparation process gave valuable findings regarding their contribution to handleabiiity. So it was possible to focus the preparation plant work on the finest products. For the flotation process, a dynamic model of the feed circuit was developed and implemented into the PLC system of a preparation plant. So the control of the flotation process was greatly improved and expensive measuring equipment was replaced by cheap software. Work on particle size sensors was done with different methods for different ranges of particle size. In the measurement of particle size above 1 mm by image analysis, a breakthrough on the segmentation of individual objects (particles) in an image was reached through the application of multi-flash imaging (MFI), here in the form of multi-wavelength imaging (MWI). Electrostatic measurements below 1 mm were applied. The method developed in this project makes it possible to calculate a mean particle size in a power station where the feeding rate to each mill is known using the existing ABB Pf meter. Pulverised fuel was analysed through laser scattering. The objective of the work was to validate an online coal fineness device at full scale in power plants. This objective was reached and allows a technical and economically feasible use of the results.