Novel Techniques of Diagnostic Data Processing for Belt Conveyor Maintenance

Abstract

In the paper a new diagnostic approach for gearbox used in belt conveyors will be discussed. The purpose of the work is to provide novel view on diagnostic data processing in the context of detection of changes in condition for population of gearboxes used in belt conveyor network. The idea will be presented by examples: a data base of diagnostic features collected during last 3 years (real data from conveyors operating in mining company) will be used for illustration. The method takes advantage from recent results of research carried out by authors and other researchers related to different types of gearboxes used in mining and other machines, (i.e. belt conveyors, bucket wheel excavators, coal shearers, wind turbines and helicopters). A serious dependency between diagnostic features and operational conditions (speed/load) it is shown in mentioned works. A novel research hypothesis has been formulated that behavior of machine in bad condition is unstable and it is more visible for heavy loaded machine. It results with diagnostic data set with higher data dispersion than for healthy one. In the paper we will prove that feature load dependency and data dispersion might be a basis for novel approach for condition monitoring of gearboxes used in belt conveyors. An advantage of such approach is its simplicity and strong physical background.