Research On The Deformation Monitoring And Forecast And Prediction Of Open-Pit Mine Slope

The problem of landslide and slope stability has been a major engineering problem withChinese characteristics since the middle and late period of the20th century owing to unique andcomplex geology conditions and the requirements of social development in China. The stabilityof open-pit slope as large artificial slope related to production and people’s lives and propertysecurity directly. Thus, monitoring open-pit slope deformation, analysis and predict thedeformation according to monitoring results is an important technology work. The deeply studyon it has important theoretical significance and application value.Open-pit slope deformation is effected by many factors and it is a complex nonlinearprocess. BP neural network is nonlinear intelligence algorithm with the characteristics of strongself-organizing ability, adaptive generalization ability and nonlinear approximation ability. Thus,it was applied widely to slope prediction. The BP neural network limitations, such as slowconvergence rate and being easy to fall into the local maximum restrict further expansion of theprediction method. To improve the performance of BP neural network, some measures should betaken. The open-pit slope time forecast by single criterion has low forecast precision and it iseasy to misjudge. Thus, to improve open-pit mine slope forecasting precision and reliability, themethod with various criterions is necessary.The main content of this paper study include Analysis the BP neural network algorithm anduse artificial fish swarm algorithm (AFSA)to optimize the initialized weights and threshold andthe structure of BP neural network algorithm. AFSA is one algorithm which has strong globaloptimization ability. At the same time establish GPS deformation monitoring system, Monitoringsurface displacement and internal force changes of rock mass of the slope. After that use the BPneural network model which has been optimized by AFSA predicting slope displacement. Anduse several criterions comprehensive forecast model based on Fuzzy matter-element theory toforecast the time of landslide. Above all is the Prediction and forecast model system of open-pitmine slope. With the use of this system by a real open-pit mine slope the result show thatAFSA-BP model have higher speed and precision than classical BP model. And the forecastmodel can also get the right result of landslide. This system is fit to pen-pit mine slope.