Highway Slope Stability Assessment Based On The Fuzzy-Neural Network

Due to the diversity and complexity of various factors, topography,geomorphology,rocks performance,geological structure, weather and hydrogeology, thephenomenon of highway slope landslide is relatively common. According to theChinese geological environment monitoring institute,the number of national geologicaldisaster is10710in total,and the number of landslide is8495,which between themaccount for78%of national geological disaster. Due to the special and complexgeological environment, chongqing is the largest number of mountain disaster, and oneof the largest in the loss rate in china. Because of a shifting climate, the impoundmentof the Three Gorges Dam and frequent extreme climate, the derivative landslide of mainroad brought many hidden dangers to the operation safety and serious damage to roadtransportation infrastructure. Therefore, combined with the engineering practice ofchongqing highway,it is necessary that we need to select the scientific and reasonableresearch methods and make a efficient and reliable highway slope stability assessment.Relying on the scientific research project of western countries—Research onhighway hazard monitoring technology with extreme weather in Three GorgesReservoir Area (2009318000001). According to the analysis of present situation of thelandslide disaster，area engineering geology survey, the main influence factors oflandslide, this paper around the field investigation of area highway slope engineeringgeological and identify highway slope stability factors.This paper calculate index weight vector by the improved AHP method. Based onthe fuzzy theory,we found the defects of the model by calculating,in view of the defectsof the model, construct and improved a fuzzy clustering model, and make a slopestability assessment by the relation-numerical method.Based on the neural network theory,construct the BP neural network model andcalculate,we found the model can learn itself and be adaptivity and fault tolerance.By integrateding the neural network theory and fuzzy theory organically,constructing the BP neural network model, and puts forward a fuzzy-neural networkmodel,using Ridge form to blur quantitative indexes,making a comparative applicationstudy through the MATLAB software, the paper get the exact,reasonable,effective,reliable and applicative highway slope stability evaluation model.