|
|
Semantic segmentation using support vector machine classifier
Pecha, Marek ; Langford, Z. ; Horák, David ; Tran Mills, R.
This paper deals with wildfire identification in the Alaska regions as a semantic segmentation task using support vector machine classifiers. Instead of colour information represented by means of BGR channels, we proceed with a normalized reflectance over 152 days so that such time series is assigned to each pixel. We compare models associated with $\mathcal{l}1$-loss and $\mathcal{l}2$-loss functions and stopping criteria based on a projected gradient and duality gap in the presented benchmarks.
|
|
|
Determination of the initial stress tensor from deformation of underground opening-theoretical background and applications
Malík, Josef ; Kolcun, Alexej
In this paper a method for the detection of initial stress tensor is proposed. The method is based on measuring distances between some pairs of points located on the wall of underground opening in the excavation process. This methods is based on the solution of eighteen auxiliary problems in the theory of elasticity with force boundary conditions. The optimal location of the pairs of points on the wall of underground work is studied. The pairs must be located so that the condition number of a certain matrix has the minimal value, which guarantees a reliable estimation of initial stress tensor.
|
|
|
Numerical realization of the Bayesian inversion accelerated using surrogate models
Bérešová, Simona
The Bayesian inversion is a natural approach to the solution of inverse problems based on uncertain observed data. The result of such an inverse problem is the posterior distribution of unknown parameters. This paper deals with the numerical realization of the Bayesian inversion focusing on problems governed by computationally expensive forward models such as numerical solutions of partial differential equations. Samples from the posterior distribution are generated using the Markov chain Monte Carlo (MCMC) methods accelerated with surrogate models. A surrogate model is understood as an approximation of the forward model which should be computationally much cheaper. The target distribution is not fully replaced by its approximation. Therefore, samples from the exact posterior distribution are provided. In addition, non-intrusive surrogate models can be updated during the sampling process resulting in an adaptive MCMC method. The use of the surrogate models significantly reduces the number of evaluations of the forward model needed for a reliable description of the posterior distribution. Described sampling procedures are implemented in the form of a Python package.
|
|
|
Reduced basis solver for stochastic Galerkin formulation of Darcy flow with uncertain material parameters
Béreš, Michal
In this contribution, we present a solution to the stochastic Galerkin (SG) matrix equations coming from the Darcy flow problem with uncertain material coefficients in the separable form. The SG system of equations is kept in the compressed tensor form and its solution is a very challenging task. Here, we present the reduced basis (RB) method as a solver which looks for a low-rank representation of the solution. The construction of the RB consists of iterative expanding of the basis using Monte Carlo sampling. We discuss the setting of the sampling procedure and an efficient solution of multiple similar systems emerging during the sampling procedure using deflation. We conclude with a demonstration of the use of SG solution for forward uncertainty quantification.
|
|
|
Geological and geotechnical characterisation of the mining environment - URF Bukov II - Interim report II
Bukovská, Z. ; Rukavičková, L. ; Chabr, T. ; Morávek, R. ; Levý, O. ; Sosna, K. ; Souček, Kamil ; Vavro, Martin ; Zelinková, T. ; Dobeš, P. ; Švagera, O. ; Kryl, J. ; Soejono, I. ; Řihošek, J. ; Holeček, J. ; Hanák, J. ; Čermák, F. ; Kašpar, R. ; Mareček, L. ; Nedvěd, J. ; Vavro, Leona ; Waclawik, Petr ; Staš, Lubomír ; Myška, O.
This report describes the work carried out in the second year of the public contract Geological and geotechnical characterisation of the rock environment – PVP Bukov II. The\ncharacterisation and descriptive works are directly related to the excavation of the premises for the future underground laboratory. These include in particular: geological and geotechnical documentation of the faces and walls of the workings, petrographic, mineralogical and structural geological documentation, hydrogeological characterisation of the environment, as well as determination of the physical and mechanical properties of the rock mass in situ and in the laboratory, geophysical characterisation and monitoring of the seismic effects of blasting and hydrogeological monitoring. An important part of this report is preparation of the classification system used to describe the rock mass. The report also includes a brief overview of the work planned for the third year of the project.
|
|
| |
|
|
Research of geotechnical methods – Stage research report – Stage C
Staš, Lubomír ; Waclawik, Petr ; Souček, Kamil ; Vavro, Martin ; Kukutsch, Radovan ; Schuchová, Kristýna ; Gong, Libin ; Kubina, Lukáš ; Zajícová, Vendula ; Nohejl, Zdeněk ; Rutar, Tomáš
The main goal of project No. FV20294 (GEOSTAB) was the verification of measured geophysical quantities with data obtained by various geotechnical methods. For this reason, stage C - Research of geotechnical methods was solved in parallel and in close cooperation with stage B - Research of geophysical methods. The geotechnical experiments in situ were carried out in the same localities where geophysical methods were applied. For this purpose, several underground mining workings located in different lithological and geomechanical conditions were selected. Experimental gallery Josef was selected in volcanic rocks, the metamorphic rock mass was represented by underground workings of the former deep uranium mine Rožná and the sedimentary environment was examined in the old slate mine Modrá štola. The stage research report summarizes the results of geotechnical, laboratory and numerical methods from all the mentioned localities.
|
|
|
Analysis EDZ using verified numerical models (FLAC3D)
Staš, Lubomír ; Waclawik, Petr ; Gong, Libin
Based on the obtained geotechnical and laboratory data, the extent and character of Excavation Damage Zone (EDZ) was simulated using static and dynamic models in the commercial software FLAC3D. The commercial program Fast Lagrangian Analysis of Continua in 3 Dimensions (FLAC3D) (Itasca Consulting Group 2012) was adopted to analyse the EDZ of Rozna Mine roadway. The analysed profile is located on the roadway No. Z3-XX in an environment formed mainly by migmatized biotite to amphibole-biotite pararules. The dynamic mode simulate using of blasting method. The research report summarize results of numerical simulation of EDZ in high anisotropic rock mass.
|
|
| |
|
|
3D laser scanning of underground workings in various conditions
Staš, Lubomír ; Kukutsch, Radovan ; Waclawik, Petr ; Gong, Libin ; Schuchová, Kristýna
3D laser scanning of mining workings were implemented at three locations with different conditions. The roadways in volcanic rocks at a depth ca 50 m below the surface at the Experimental gallery Josef, selected part of the Z3-XX roadway located in a metamorphic rock mass at a depth about 950 m below the surface at the Rožná mine and the closed slate mine Modrá štola in sedimentary rock mass located just below the surface, were scanned and interpreted in detail. 3D scans of mining workings were mainly used as input data to the GENIE software for processing geophysical measurements in the 3D dimension. This software tool was developed by the staff of the Technical University of Liberec and Geotest a.s.
|
guest ::
RSS feed.