Sistema de monitoreo para la detección del estado de vías aplicando técnicas de instrumentación basadas en técnicas de minería de datos a partir de variables electromecánicas
Ibáñez Noriega, Iván | 2018-10-05
Coal mining is currently going through a period of recession which is leading senior
management to look for new tools to optimize processes and stay active in mining market.
One of the action points considered is the management of equipment maintenance due to the
poor condition of the roads, making necessary the implementation of effective methodologies
to make adjustments to the optimal inventory point. Among the techniques that consider data
mining are the so-called SOM acronym in English for "Self-Organizing Map", which allows
the user to train an algorithm and make a map on the associated input variables through
patterns, to generate a response based on these in order to clear doubts or establish reference
points. This research considered as input variables the databases with selected variables of
electromechanical relevance of a fleet of mining trucks, where the neural network algorithm
will associate these data and find patterns that facilitate to determine through a mathematical
analysis the condition of roads with a high degree of reliability, this information is a key
piece for planners to establish the necessary maintenance routines and determine the optimal
point of dispatch of resources of support equipment for repair.