Study On The Influence Of Bank Collapse On Qiaomeng Highway In The Area Of Baihetan Reservoir

The reservoir area of Baihetan hydropower station is located in the lower reaches of the Jinsha River,the Dam site of which is 45 km away from the upper Qiaojia County.It is located in the southeastern margin of the Qinghai-Tibet Plateau,belonging to theAlpine and plateau geomorphic units of southwestern part of Sichaun Province and the northeastern part of Yunnan Province,which is a part of Hengduan Mountains.The area is roughly bounded by Xiaojiang,and the west is Kangdian ground axis area,which is influenced by the Xiaojiang fault and Themuhe fault.The maximum water level of Baihetan Hydropower Station is 825.0m,the limit water level for flood control is 785.0m,the dead water level is 765.0m,and the total reservoir capacity is 20.627 billion m~3.Baihetan reservoir is operated with the water level elevation between 765.0m and 825.0m every year.The high way between Qiaojia and Menggu is intended to be built within the water level variation with the length of 75.323 km.The bank slope will inevitably lead to bank reconstruction,resulting in bank collapse,traffic disruption,threatening the safe operation of reservoir area and Qiaomeng high-speed construction.According to the factors such as the position of highway line,geological environment and water level fluctuation,the K0-K51 section is the main influence area of the bank referred to the study area.It is necessary to carry out the research based on bank collapse model,forecasting method,forecasting parameters,improving the prediction accuracy of bank collapse and engineering practicability.Through the optimized model of the bank collapse width,combined with the proposed highway conditions for evaluation,it is of great theoretical and practical significance to guide the construction of Qiao Meng highway.Based on the engineering geological conditions of the reservoir area,this paper combined the example of Zipingpu,Tianshengqiao and the Three Gorges,which have the similar geological background and the resvoir type.The structure,stability and water storage of the study area.In this paper,the improved in-situ borehole shear strength parameters and fractal theory are used to predict the characteristic parameters.Based on the combination of traditional classification methods and engineering practice,this paper establishes the attribute reduction with Rough Set Theory,considers the BP neural network prediction model of bank collapse model,and predicts the bank collapse width.Combined with the forecast results,the impact of the collapse for is evaluated.The main research results are as follows:(1)The bank slope in the study area is mainly divided into soil bank slope and rock bank slope.The soil bank slope is mainly composed of the Quaternary loose deposit.Taking the engineering geological rock group as the basic unit,considering the factors such as bank structure type,topography and topography,the whole bank is divided into 12 sections and 21 subareas,among which there are 5 sections of 11 rocky sections,about 17.98 km,accounting for about 35.20%;soil bank is divided into 7 segments and 10 subareas,about 33.02 km,accounting for about 64.80%.(2)The stability of each bank is classified according to the factors such as bank structure,distribution characteristics of rock and soil,topography and geologic phenomena.There are 4.8 km in class I,37.56 km in Class II and 8.64 km in class III.The length of stable and basically stable bank slope is 42.65 km accounting for 83.06%,and the length of poor stable and unstable bank slope is 12.61 km,accounting for 16.94%.The stability of the whole Qiao Meng bank slope is better.(3)According to the characteristics of investigation,stability status and similar types and geological conditions of the mountain river bank collapse features,combined with the current development of the bank slope,the bank collapse model of the research area is divided into erosion type,According to 46 measurement and experimental studies,it is found that the proportion of slip-type collapse bank is about 52%,and the erosion type and collapse type are 24% and 13% respectively.The collapse of the rocky slope section is more stable,with a small scale superficial weathering layer collapse,out of block,accounting for 11%.(4)The mechanical strength parameters of natural and saturated soils in different rock and soil were determined by in-situ borehole shear test,and the fractal characteristics of different loose bulk particles and abrasive erosion angles were established by fractional test.The fractal dimension of the bank material composition is different,and the fractal dimension is proportional to the volume of the fine particles.(5)Based on the background data of the reservoirs with similar geological conditions,the slope roughness,slope characteristics,abrasive erosion angle,cohesive soil content,cohesion force are obtained based on the RS rough set theory.The internal friction angle is the main influencing factor of bank collapse.The non-linear mapping relationship between reservoir bank collapse model and reservoir bank collapse prediction property is studied.The BP neural network model with double nested function considering the collapse model based on reduced attribute is established,which is well optimiazed at the end.