Control over crushing and screening with the help of visiometric analysis of ore

Abstract

The developed method of visiometric analysis of particle size distribution of crushed ore uses the procedures of histogram image processing, thresholding of halftone areas, spatial image filtering, contour extraction, visualization of segmentation results, and calculation of particle size distribution histograms using the gamma function. This method allows to determine quite accurately the mass fraction of the specified sizes in crushed ore. Relationships were determined between the main output parameters of the process — i.e. the output of the final product and the screening efficiency — and the main input parameters, such as the crusher capacity, the open side setting, and the energy consumed for crushing. The obtained dependences are described by second-order multiparametric regression equations, which were used to establish the coefficients of proportionality between the deviations of the control signal (the open side setting and the throughput) and the deviations of the measured parameters (the output of the oversize product, screening efficiency, energy consumption). The developed system to control crushing and screening processes comprises scales and visiometric ore size analyzers installed on the conveyors, a meter that measures the power draw of the crusher motor, a control unit, an open side setting regulator and an ore flow regulator. The following criteria were used for optimized control of the crushing and screening processes: the screening efficiency for the –12+10 mm product and the yield of +2–10 mm size in the crushed ore. A fuzzy logic technique was used for automatic control. This technique determines the direction and the change of the controlled parameter as a function of the changing values of the measured process parameters. The results of industrial tests show that the developed optimization technique allows to obtain a crushed product of a given size (95% of the –12 mm product), increase the throughput by 5.7% and reduce the total energy consumed by the ore conditioning circuit by 2.8%.