Na tentativa de contornar os problemas ainda existentes para a detecção e identificação de erros grosseiros (EGs) no processo de estimação de estado em sistemas elétricos de potência (EESEP), realiza-se, neste trabalho, uma análise da formulação dos estimadores aplicados a sistemas elétricos de potência, em especial, o de mínimos quadrados ponderados, tendo em vista evidenciar as limitações dos mesmos para o tratamento de EGs. Em razão da dificuldade de detectar EGs em medidas pontos de alavancamento, foram também analisadas as metodologias desenvolvidas para identificação de medidas pontos de alavancamento. Através da formulação do processo de EESEP como um problema de álgebra linear, demonstra-se o porquê da impossibilidade de detectar EGs em determinadas medidas redundantes, sendo proposto, na seqüência, um método para identificação de medidas pontos de alavancamento. Para reduzir os efeitos maléficos dessas medidas no processo de EESEP verifica-se a possibilidade de aplicar outras técnicas estatísticas para o processamento de EGs, bem como técnicas para obtenção de uma matriz de ponderação adequada. / To overcome the problems still existent for gross errors (GEs) detection and identification in the process of power system state estimation (PSSE), the formulations of the estimators applied to power systems are analyzed, specially, the formulation of the weighted squares estimator. These analyses were performed to show the limitations of these estimators for GEs processing. As leverage points (LP) represent a problem for GEs processing, methodologies for LP identification were also verified. By means of the linear formulation of the PSSE process, the reason for the impossibility of GEs detection in some redundant measurements is shown and a method for LP identification is proposed. To minimize the bad effects of the LP to the PSSE process, the possibility of applying other statistic techniques for GEs processing, as well as techniques to estimate an weighting matrix are also analyzed.

Name of the thesis is "Effects of the Teaching Czech as a Second Language" and in its first part presents theoretical introduction to the field of study, empirical researches of student's productive skills (Czech as a foreign language) follow (speaking, writing). First part is focused on theoretical knowledges of second language acquisition, options and approaches to exploration of language's errors and approach and viewing of errors and mistakes in general, especially in foreign researches and literature. Bigger attention is paid to selected analyzes, those are basic instruments for making of practical part of the thesis. It is especially error analysis (EA) and analysis of fluency and accuracy. Because the thesis's topic is concerned to effects of teaching and affecting factors, there are defined terms of pedagogy and psychology in theoretical part. In practical part there is presented empirical execution of mentioned analyzes, its comparison, evaluation and making conclusions in relation to appointed variables, or rather determinants, which enter the education process. Beside these instruments the research uses questionnaires and evaluation sheets for (self)reflection of student and lecturer. The aim of the thesis firstly is mapping of student's interlanguage progress thanks to case study, and...