A hierarchical procedure for adjustment and material balancing of mineral processes data

D. Hodouin,M.D. Everell
DOI: https://doi.org/10.1016/0301-7516(80)90002-2
1980-09-01
International Journal of Mineral Processing
Abstract:In order to evaluate and model mineral processes, it is very useful to obtain consistent data in order to satisfy mass conservation. Since collected data are often redundant and erroneous due to natural disturbances, sampling errors, unreliable instrument readouts and laboratory analyses inaccuracies, they must be adjusted. The proposed procedure to minimize the weighted sum of squared residuals consists in a supervisory program working on a set of independent variables necessary to define overall ore flow rates. Within this algorithm, the variables are separated into two sets: the microscopic ones (particle size distributions, chemical assays...) and the macroscopic ones (flow rates, mass fraction of solids in pulp...), each set corresponding to a part of the sum of squares criterion to be minimized. The minimum of the microscopic part of this function can be found directly since the constraints are linear and the function quadratic. The macroscopic part is minimized using the standard Powell algorithm which is also used as supervisory algorithm. The material balance constraints are written using matrix formalism and flow-network description of the circuit. Three industrial application examples are given: a flotation circuit for which ten chemical elements were analyzed, a grinding circuit closed by DSM screens and a three-stages grinding circuit with rod, ball mills and hydrocyclones.
What problem does this paper attempt to address?