Application of Artificial Intelligence in Recultivation

Abstract

Developing a recultivating technology for existing refuses (spoil-bank) and reservoirs of slurry, and the elaboration of a monitoring system, have become one of the most significant tasks in Pécs and its neighbourhood because of previous mining activities. Sampling haens in well-defined (relatively few) places during the collection of data. For creating a digital urface model, drawing contour lines and for data evaluation, not only the information of places where measuring happened are needed, but we have to include the data of places where sample collection did not or could not take place. (In order or be able to create an exact surface model with acceptable resolution, we need data from numerous places.) Since we do not possess data from each point of a given area (it is impossible to collect and store data from infinite places), we have to use interpolation. To do that – in technical literature – there are several solutions: statistical functions, 3D evolutional algorithm, neural network, Fuzzy algorithm and fractals. We examined a method that is new and rarely used in special informatics to process data. It is the neural network (NN).