Application Of Genetic Algorithm And Artificial Neural Network In The Evaluation Of Slope Stability

Landslide frequently occurs in China that has threatened the people's life and property and the sustainable development of economy and society. With the continuous development of our country's modernization, many slope projects and more and more high and steep slopes have appeared. Therefore, analyzing and evaluating the slope stability in an accurate and reliable way will play a significant role in predicting and forecasting the landslide.Regular evaluation methods of slope stability and their advantages and disadvantages are summarized and epitomized in this paper, the development trend of the evaluation methods are also presented. In order to solve the multi-factor problem of evaluation of slope stability, the projection pursuit classification model(PPCE)and projection pursuit grade evaluation model(PPGE) are suggested and applied, which are founded on real coding based accelerating genetic algorithm(RAGA).The algorithm and the step are also provided. Using RAGA to construct the PPCE model and the PPGE model can optimize the course of the traditional projection pursuit method. The practical example shows that both PPCE model and PPGE model are effective and feasible. BP neural network is a complex nonlinear dynamic system. It has a strong nonlinear mapping ability and can construct the nonlinear relationship between factors and slope stability through the learning of some examples of slope. By making use of the nonlinear relationship, the slope stability with high accuracy can be evaluated.