Theory And Method Reasearch Of Safety Assessment Of High Rock Slope Based On Displacement Information

The stability of rock slope essentially depends upon deformation. Inquiring the deformation condition of slope in different phase is important to appraising the stability of slope. As the important tool of exactly evaluating the alteration of the slope status, safe monitoring has been practiced in many high slope projects. Due to the complexity of slope, the research on the monitoring theory is not systematic, and the analysis of monitoring data is not profound. On the basis of referring on large number of correlation literature, this dissertation is focused on the thorough analysis of monitoring displacement information of major slope engineering, and presents a trial study on the theory of monitor control. The main contents and research results are as follows:(1) The influence of error in the observation to the slope behavior appreciation and parameter back analysis is analyzed. According to the characteristics of slope project, the rule of displacement monitoring points arrangement is put forward. In practice the arrangement can be directed through sensitivity analysis and numerical simulation. Example study shows that the abnormal deformation can be detected in time if monitoring points are located on the position sensitive to displacement variation. And the reliability of back analysis can be improved if the monitoring points are set on the location that sensitive to parameter variation.(2) The error in the monitoring data may influence the accuracy of data analysis, so it is necessary to recognize the rough error and pick out. However, the tradition methods dealing with error is not suitable to the data of slope displacement. So the uncertain filtering method in the uncertainty mathematics is introduced, and the improved filtering method is utilized in the data treatment. Case study indicates that rough error and exception deformation value can be identified by this method.(3) The deformation of slope is composed of two parts: the rheologic feature of material and the influence of external condition. Based on the analysis of large amount of surface monitoring data, the statistic model describing the deformation character of rock slope is put forward. The abnormal deformation value in the monitoring data is analyzed by the fiducial interval method and little probability method. The trend in residual error is fitted by periodic function, and the model is modified.(4) The deformation control criterion is researched. The deformation data is composed of three parts: stable deformation value, theological deformation component and measure error. Threshold value of standard deviation in different phase is deduced, and the phase of slope deformation can be judged by computing the standard deviation of data series. When the slope is in remarkable deformation stage, the increasing trend of displacement mean values can be evaluated by statistical quantity(r), and divergence of data can be evaluated by coefficient of variation. If the value of statistical quantity(r) is high and value of coefficient of variation is small, the slope deformation is regular. On the basis of analyzing displacement quantity, the displacement direction and speed rate is considered comprehensively. And the surface displacement data of practical engineering is analyzed through the multi-parameter index.(5) The limit state of slope is researched by means of safe factor. Safe factors can be obtained by strength reduce method and limit equilibrium method, depending on the slope type. Due to the feature that slope deformation is big before collapse, the safe factor and the position of dangerous failure plane is computed through FLAC code. Besides, the local safe factor can be defined by plotting the curve between the displacement sensitivity and reduction coefficient, thus the overall stability of slope can be evaluated. On account of the problem that the parameters in the limit equilibrium method is hard to determine, the blind data is introduced to deal with the uncertainty in parameters, and improvement has been made to this aspect.(6) Different forecasting method is used according to different size of monitoring data. When monitoring data sequence is long, the relation between observation value is described with polynomial, and pso algorithm is used in model optimization. When the sample data is small, the improved grey model is used in the forecasting of inequality time interval data. The original data can be handled with weighting method in order to increase the precision of forecasting.(7) The forcasting of slope collapse is researched via deformation response ration. Deformation of slope can be seen as response to generalized load. The failure time can be judged through computing response ratio of deformation to load in different time. The monitoring point where deformation is in safe state is chosen as base point, and the displacement change of base point is treated as the change of external load. Make comparison between the displacement increments of other monitoring points with that of base point, and calculate the response ration. Then according to the change of response ration the failure forecasting can be performated.(8) The comprehensive assessment system is established on the basis of monitoring information on site, and the rule of point rating about bottom index is brought forth. In order to get the appraisal that reflect the diversity of the slope condition, attribute recognition theory is introduced in research. The result of attribute cognition is qualitative, so it is not easy to make quantitative comparison with the slope conditions in different times. The catastrophe theory is used in the comprehensive assessment and the specialty of catastrophe theory in the multi-rule assessment is utilized to avoid the subjectivity of artificial weight. After standard quantizing of bottom assessment index, the ultimate catastrophe value is obtained through quantitative computing by the normalization formula of different catastrophe model. Thus the development of stable status of the slope can be obtained by comparing ultimate catastrophe values of different times.