Condition Monitoring for LHD Machines Operating in Underground Mine—Analysis of Long-Term Diagnostic Data

Abstract

Load–haul–dump (LHD) machines are key assets in horizontal transportation in the underground mine. They perform ore haulage from mining faces to the nearest belt conveyors, dumping the material at the specific locations. Their operation can be described as cyclic. For maintenance staff of those machines, effectiveness-related demands are the greatest challenge, especially taking into consideration harsh and time-varying environmental conditions. The analysis of long-term data recorded on such machines can provide information about the changes in technical condition of the machine. Moreover, such observations can allow to track slowly progressing changes in technical condition that are effectively impossible to asses given only short-term measurement data. Besides the context of analyzing such data, it can be used as large-scale training dataset carrying a lot of useful information for the future development of diagnostic procedures. In this paper, authors propose statistics-based methodology for the analysis of long-term observations of diagnostic data recorded on LHD machines. Fusing information contained in various types of diagnostic variables (i.e., temperatures, pressures, operational parameters like engine rotational speed, torque, etc.) can allow to unravel underlying degradation processes occurring in the machine, with greatest focus on drive-related components.