Tailings Dam Stability Reliability Analysis Based On Neural Networks And Influencing Factors Research

Tailings installations are the important part of mine production facilities,and the safe operation tailings dam directly related to people's lives and property downstream, as well as the ecological environment around the reservoir area. Therefore, the reaearch on stability of the dam not only has academic value,but also has both economic and social benefits.In the tailings dam project analysis process, traditional evaluation is to use the safety factor method. It does not take the variability of the soil into consideration,but considering the soil as a kind of 'average' material of the'homogeneous' materials.so,it set some parameter to various design condions and a variety of indicators. Furthermore,selecting a certain mode of calculation to be calculated.On the contrary,the unknown, uncertain factors are classfied into a single safety factor. In that case, the results are obtained is clear.So, the trational method is acceptable. Howerer,This traditional methord ignores the nature of the geotechnical parameter uncertainty, and is not consistent with the fact. In recent years, There is a growing recognition in the tailing dam problems of uncertainty.At the same time, it is an inevitable trendency of development to introduce reliability analysis method into the tailing dam project.Different from the traditional safety factor method, the reliability analysis of neural networks in the tailings dam stability takes soil parameters'random, variability and non-linear relationship into consideration, to use the probabilistic approach to quantitatively consider the actual existence of these uncertainly factors.So,it can be more objective,quantitative reflected the actual safety of dam.This paper puts forward the reliability of neural network analysis method.It is a combination of both the BP neural network method and Monte Carlo method. reliability Analysis of the stability of the tailings dam, and a preliminary discussion of the tailings dam reliability standards. The reliability of neural network analysis method uses BP neural network mapping ability of strong nonlinear function,The tailings dam stability factors as inputed neurons,as well as safety factor as the outputed neurons, In the MATLAB platform to build a dam safety factor predicted BP neural network model, combined with the Monte Carlo simulation method, to solve the reliability index problem.at the same time, this paper also uses Grey Incidence Analysis Model to analysis various cases effect on the stability of the dam.Through the stability of the tailing dam and the inflence factors of gray incidence analysis to idenrity the dominant factor of the dam stability. Compared to the traditional analysis method which take only single-factor effect into consideration. we can obtain a reasonable system of internal rules of Luwanqing tailings dam.So as to provide basis for the manage and optimize design of it.