MODELOS MEF 4D, REDES NEURONALES Y FUZZY LOGIC DE LA MAQUINARIA Y EQUIPOS EMPLEADOS EN OBRAS SUBTERRÁNEAS

Abstract

The results of this paper are derived from an Investigation Project financed by Technological and Industrial Development Center of the Ministry of Industry and Energy (University of Oviedo – a private company Carbonar S.A.). Graphical Engineering, Neural network, FEM3D, Fuzzy-Logic and FEM4D have been applied to mining and general underground works in this project. A deep measurement campaign has been widely developed by model adjustment, data being processed through SPSS, Microstation, SiteWorrk and Surfer. A Neural network has been created permitting us to obtain the loading values over roof support, the distance between roof weighting and overloading frequency. The latter has been obtained by the layer strength, the roof rock stratum and the stratum quality index. Being carried out the roof support unit FEM3D models (piece by piece and assembling), FEM4D models were used introducing the advance movement variables through Fuzzy- Logic models and the neural network values achieved. These are powered tools that can be used to presuppose knowledge of the stress-deformation variation at any real time, which are operating over machinery and equipment in underground activities. Keywords: Underground works, Underground mining, FEM, Neural network, Fuzzy Logic, Longwall, Roof support