TASK Quarterly   Scientific Bulletin of the Centre of Informatics - Tricity Academic Supercomputer & networK   ISSN 1428-6394

4/2010


Contents:



Abstracts:

  • M.Szymaniak, A.Gardzilewicz, Modernisation of Steam Turbine Construction at Regenerative Steam Extraction Point

    The paper presents a methodology which has led to improvement of the turbine stage performance in the regenerative steam extraction chamber area by introducing a ring which collects the leakage flow from the gaps over unshrouded rotor blade tips and directs it to the extraction chamber. A comparison is made between the results obtained for the initial design (without a ring) and the design with the ring mounted. The results obtained for the construction were calibrated using the data measured on a real turbine in a Polish power plant.

  • M.Kmiotek, A.Kucaba-Pietal, Preliminary Calculations of Flow in Channel with Triangular and Rectangular Obstacle

    The paper presents the results of preliminary numerical calculations of a fluid flow in a channel with an obstacle. The flow problem was solved with an application of the finite elements method. Various geometries of obstacles were considered.

  • M.Cudny, P.Popielski, Analysis of Excavation-induced Deformation with Different Soil Models

    The monitored case of deep excavation works located in the centre of Warsaw is described and back analysed as a boundary value problem with the finite element method. The excavation was carried out in over-consolidated clayey layers under the support of braced diaphragm walls. Accurate simulation of such soil-structure interaction problems requires advanced soil constitutive models - especially for the pre-failure range of small strains. On the other hand, such models should not be very complex and their material parameters should be relatively simple to obtain from laboratory and in situ surveys. By means of a case study, this paper examines several simple elasto-plastic constitutive models for the simulation of the behaviour of soil layers. The influence of such characteristics of the soil behaviour, as anisotropy and non-linearity of stiffness, is studied in the paper. The discussion concerns the problems related to the application of the models for a practical example. The comparison is focused on the obtained and measured displacements of the neighbouring building and diaphragm walls.

  • A.Boltuc, E.Zieniuk, Application of PIES and Rectangular Bezier Surfaces to Complex and Non-homogeneous Problems of Elasticity with Body Forces

    This paper presents a variety of applications of an effective way to solve boundary value problems of 2D elasticity with body forces. An overview of the approach is presented, its numerical implementation, as well as a number of applications, ranging from problems defined on elementary shapes to complex problems, e.g. with non-homogeneous material. The results obtained by the parametric integral equation system (PIES) were compared with the analytical and numerical solutions obtained by other computer methods, confirming the effectiveness of the method and its applicability to a variety of problems.

  • M.Wroniszewska, J.Dziedzic, Voice Command Recognition Using Hybrid Genetic Algorithm

    Speech recognition is a process of converting the acoustic signal into a set of words, whereas voice command recognition consists in the correct identification of voice commands, usually single words. Voice command recognition systems are widely used in the military, control systems, electronic devices, such as cellular phones, or by people with disabilities (e.g., for controlling a wheelchair or operating a computer system).

    This paper describes the construction of a model for a voice command recognition system based on the combination of genetic algorithms (GAs) and K-nearest neighbour classifier (KNN).

    The model consists of two parts. The first one concerns the creation of feature patterns from spoken words. This is done by means of the discrete Fourier transform and frequency analysis. The second part constitutes the essence of the model, namely the design of the supervised learning and classification system. The technique used for the classification task is based on the simplest classifier - K-nearest neighbour algorithm. GAs, which have been demonstrated as a good optimization and machine learning technique, are applied to the feature extraction process for the pattern vectors. The purpose and main interest of this work is to adapt such a hybrid approach to the task of voice command recognition, develop an implementation and to assess its performance.

    The complete model of the system was implemented in the C++ language, the implementation was subsequently used to determine the relevant parameters of the method and to improve the approach in order to obtain the desired accuracy. Different variants of GAs were surveyed in this project and the influence of particular operators was verified in terms of the classification success rate.

    The main finding from the performed numerical experiments indicates the necessity of using genetic algorithms for the learning process. In consequence, a highly accurate recognition system was obtained, providing 94.2% correctly classified patterns. The hybrid GA/KNN approach constituted a significant improvement over the simple KNN classifier. Moreover, the training time required for the GA to learn the given set of words was found to be on a level that is acceptable for the efficient functioning of the voice command recognition system.

  • A.Piorkowski, M.Zupnik, Loop Optimization in Managed Code Environments with Expressions Evaluated Only Once

    This paper is concerned with optimizing code execution in virtual machine environments. Code motion is one of the many optimization techniques. We considered a special case of optimization - a loop containing expressions that can be effectively evaluated once. A solution for this case is discussed and two algorithms are proposed. Experimental results for Java VM, MS .NET and Mono are shown here in order to assess the performance of the proposed algorithms.

  • L.Pasini, S.Feliziani, Graphical Application for the Construction of a Simulator of Urban Vehicular Traffic Flows

    One of the most fundamental problems encountered during the construction of a simulator of urban vehicular traffic flows concerns the method of gathering the data describing the real system. In this paper, we present a graphical application developed for AutoCAD. This application allows to process the images of a vehicular traffic system in AutoCAD. During the graphical processing performed by means of our application, a file describing the system is generated. This file is subsequently read by a procedure which generates the simulator of the system. The images of the traffic system can be acquired by means of several applications available on the Internet, such as Google Maps, Google Earth, Bing, etc. Performing image processing in the context of AutoCAD allows to consistently handle distance measurements in the system, and, in general, to address the problems related to the scale ratios of an image.