|
|
Analysis of the time evolution of selected parameters of fires for the period 1977-2016 on the territory of the present Czech Republic
HRNČÍŘ, Miroslav
The thesis deals with the fulfilment of the following objectives: Objective 1: System Analysis of emergency events associated with fire protection, the causes of fires and the link to the system environment. Objective 2: Examination of the homogeneity of the selected time period in terms of the survey of selected parameters (number of fires, damage, number of injured people, number of people killed). Objective 3: Examination of paired dependence of selected parameters (number of fires, damage, number of injured people, number of people killed). Objective 1 was fulfilled on the basis of the system analysis performed and evidenced on individual chapters of 1.1 to 1.7 of the theoretical part, which describe the examined problem as a system with structure and with links to the system surroundings and the system superior. Compliance with Objectives 2 and 3 was associated with the verification of the H1 hypothesis (and its sub-hypotheses H11, H12) and H2 hypothesis. The thesis explored the following hypotheses: Hypothesis H1: The selected fire parameters in the monitored period are statistically dependent. Due to the system analysis, this hypothesis was divided into two partial hypotheses: Hypothesis H11: Statistical dependence will be demonstrated for selected parameters in terms of time development. Hypothesis H12: Statistical dependence will be demonstrated for selected parameters from the viewpoint of paired bonds between the parameters. Hypothesis H2: The theoretical distribution of selected fire parameters is the normal distribution. The H11 hypothesis was accepted. In terms of time development, determination coefficients from 0.4234 to 0.8271 were found. Correlation coefficients from 0.65 to 0.91 corresponded to this. The values of all the examined parameters, i.e. "number of fires", "damage", "number of injured people", "number of people killed" increased with the time. The H12 hypothesis was accepted. The examined paired relationships correlated with values of 0.72, 0.80, 0.69 and 0.67. The global H1 hypothesis was adopted. Contrary to the H1 hypothesis, the H2 hypothesis was basically rejected - normality was proved only in one parameter ("number of people killed"), in the other three cases it would be necessary to look for another theoretical division. The rejection of the H2 hypothesis was supported by both acceptance of the H1 hypothesis and the detection of the inhomogeneity of the 1977-1991 period (some homogeneity emerged for the partial periods of 1977-1991 and 1992-2016). The benefits of the diploma thesis can be seen in theoretical part (justified application of mathematical statistics without scaling in time development, justified application of mathematical statistics with scaling in paired dependencies) as well as in the practical part (finding specific time and pair correlations). Based on the description of obtained results, the possible suggestions for follow-up work were described. To carry out partial investigation from the viewpoint of two identified partial periods with possible data homogeneity, to carry out an investigation with an extended number of parameters based on e.g. factor analysis.
|
guest ::
