关键词: 风险分析, 算法, 地震灾害, 地质条件, 模糊系统

Abstract: Risk analysis about earthquake involves many different aspects such as seismic intensity, site condition, distribution and antiseismic capability of buildings, supporting system of disaster precautions and rescue etc. However, little effort has been made to quantitatively research the influences of these factors on earthquake disaster risk so far. In this paper, the evaluation of the influence of geological factors on earthquake disaster risk is discussed in detail. There are many geological factors that influence significantly the degree of seismic hazard, such as stratum, groundwater, topography, seismogenic faults etc, among which the stratum plays an important role. The algorithm for evaluating the influence of geological factors on earthquake hazard risk is established here by considering mainly stratum, especially soft formation. The evaluating indexes are mainly lithology, buried depth and thickness of the stratum, and the lithology index is quantitated with shear wave velocity (v_s). In order to elementarily dispel the interference of earthquake magnitude and epicentral distance in the analysis of the relationship between earthquake hazard of the buildings and the site conditions, it is necessary to normalize the earthquake hazard data from different intensity background by timing a coefficient. After dividing by the maximum of the obtained data, the normalized data is simply called Relative Earthquake Disaster Coefficient of Buildings (RDC). The risk factors are very complicated, so it is extremely difficult to collect enough data and set an organic connection between the geological factors and earthquake hazard variables in statistic meanings. Therefore, the theory and methodology of fuzzy system are needed for the analysis. After giving the domains of the variables, a fuzzy relation matrix between lithology (L), thickness (T), depth (D) of the stratum and RDC can be established with Multi-dimension Information Distribution. On the basis of it, the Possibility-Probability Fuzzy Risk, which reflects better the nature of the risk, needs to be calculated but lacks an algorithm in the case of multi variables (here is 4). A universal algorithm with its corresponding definitions, formulas, and calculating steps is established in this paper, which is applicable to the similar analysis even if the input variables are different. When the values of lithology, thickness and depth of stratum samples are input into the calculation model of the Possibility-Probability Fuzzy Risk, at first the possibility-probability distribution of the combination of these stratum factors is calculated on their own variables space, while earthquake hazard risk analysis requires the evaluation of the contribution of those factors to earthquake disaster indexes in another space. Risk analysis on the basis of possibility-probability distribution of relevant factors in different spaces, however, has not been reported in the previous work. In this paper, the question is solved by means of fuzzy information transmission of risk possibility with a given formula, utilizing the fuzzy causality matrix between stratum factors and RDC. An example of how to use this method to evaluate the influence of stratum factors on earthquake disasters risk is illustrated by the data of the stratum and earthquake hazard of Xingtai large earthquake in 1966 and Haicheng large earthquake in 1975. Although lithology, thickness and depth of the stratum are taken as the basic input parameters in the above mentioned example, one may increase, decrease and change the input parameters according to practical needs and information of risk evaluation.

Key words: risk analysis, algorithm, seismic disaster, geological conditions, fuzzy system