Stanis3aw Gruszczyñski, Krzysztof Urbañski: Application of Interpolation Algorithms and Artificial Neural Networks for Chromium Contents in Soils Characterization In¿ynieria Œro- dowiska

Various ways of approach, to determine the horizontal distribution trend (tendency) of Chromium (Cr) in soil, where is high pollution by this element are analysed. Polynominal regression algorithms (I, II, III degree polynominals), interpolation algorithms (TIN, Kriging, RST), and also artificial neural networks (MLP, CANFIS, RBF, GRNN, PNN, MDN) are applied. Data from field experiments, carried out in the area of Chemical Plant in Alwernia were used. The differences between several ways of approach are presented in a graphical form, and also in some remainders distribution statistics. The soil pollution spatial distribution examinations lead to following conclusion, that in the first place is the information precision determination, and also the limit of error, through the pollution evaluation acceptance, whereas in the second place is the indication or standing out the regularity connected with the imission effect mechanism. It seems that the chromium concentration in soils variation, noticed even on short distances, makes it difficult the acceptance of interpolation method, as a method of contamination distribution evaluation. On the other hand the considerable nonlinearity makes difficult the acceptance of regression model. In these circumstances, the possibility which is worth consideration, is the modelling with the application of neuron networks, that is also hybrid solution application (for instance MDN), which gives the possibility of Cr concentration in soils variation deeper analysis.