A.A Morshedy; H Memarian
Abstract
Various interpolation and estimation tools are used to spatially model a regional variable across an area or site. This paper presents a new interpolation method, using the progressive radial basis function network and taking into account the spatial coordinates of the input data. The procedure starts ...
Read More
Various interpolation and estimation tools are used to spatially model a regional variable across an area or site. This paper presents a new interpolation method, using the progressive radial basis function network and taking into account the spatial coordinates of the input data. The procedure starts with the study of the spatial structure and anisotropy of the data, to perform interpolation and determining the radiuses and rotation angles based on the directional variography. Next, the neighborhood radius and neighboring points of each node of hidden unit are determined, using the ellipsoidal anisotropy and the covariance matrix. Then, a shape factor is computed based on half the average distance of all the neighboring sample points. The progressive kernel matrix includes the corrected kernel functions and the coordinates of the nodes in the hidden units utilized to solve the weight matrix. The interpolation was finally performed at each point of regular network (unsampled points). The steps of this interpolation algorithm were evaluated by a synthetic data set, having an irregular 3D pattern. The Cross validation between actual and estimated values have a correlation coefficient of about 0.78 and the fitted line passing through the actual and estimated values is close to 45 degrees.
A. Hossein Morshedy; H. Memarian
Abstract
Zoning is an important practice in earth sciences. In zonation, the study area is divided into separate parts and by compiling the results of these parts, a unique model is obtained. In this study, clustering methods are applied for zoning of Semilan dam site. Optimal number of clusters are measured ...
Read More
Zoning is an important practice in earth sciences. In zonation, the study area is divided into separate parts and by compiling the results of these parts, a unique model is obtained. In this study, clustering methods are applied for zoning of Semilan dam site. Optimal number of clusters are measured based on geotechnical parameters (lugeon, RQD), the importance of various dam structures and lithology indicators. By ranking of 7 clustering validity indexes, the optimum number of clusters found to be 4. In this paper, clustering was performed by faults locations and self-organizing neural network. In the former case, the study area was divided into four zones based on faults. This two dimensional zoning is independed of the third dimension (depth) and each sample belonged to a cluster. In the later case, a self-organizing map (SOM), which is a kind of neural network capable of clustering, was used. The SOM input data consists of, three dimensional parameters (X,Y,Z), geotechnical parameters (lugeon, RQD) and finally indicators of importance of various dam structures and lithology. Then, 7 input parameters were normalized between 0 to 1 and entered the network for training.The output data were allocated to four zones (clusters). For RQD spatial distribution realization, variography and anisotropy parameters for all four zones were calculated for both cases, Based on the main principal of clustering method which is maximum difference between clusters and maximum similarity between members of each cluster, performance and validation of two cases of clustering, RQD data were defined. Clustering quality index defined as sum of mean differences between two clusters divided by sum of standard deviation of clusters. Maximizing of this index is optimal solution. This study showed that clustering by SOM gives more accurate results than clustering by faults.
