| The landslide is a multiple of geological disasters. China is the world’s landslidedisaster-prone countries in the world, there are nearly a quarter of the land area may orbeing subjected to the threat of landslide disasters, the resulting personal and propertydamage and economic loss is not counting. With the rapid development of China’seconomic construction, to carry out large-scale geotechnical engineering construction infull swing, the Landslide rock mass displacement sequence predicted more and moreattention has been paid. Ability to accurately predict the forecast landslide deformation,so as to provide the basis for timely adjustment of the construction program and take thenecessary emergency response, it has important practical significance.In this paper, on the basis of the Markov chain model, the use of fuzzy thinkingand combination forecasting method to predict the deformation of the landslidedisplacement, mainly following results were obtained:(1) Fuzzy thinking Markov chain model is applied to the landslide forecast andFuzzy Grouping parameter to Markov significance level preferred by MATLABprogramming in the average deformation the Liangshuijing landslide order predictionuse is preferably a single interval accounted parameters to achieve a better predictionresults, the use of fuzzy thinking level features worthy to accurately forecast value, theresult and the true value comparison error is less than1.2mm, the average error does notexceed0.5mm.(2) The first gray-fuzzy weight Markov chain model is applied to the landslidedisplacement deformation prediction, showing very significant Markov and achievedsatisfactory prediction. The results are compared to the gray theory predictssignificantly improve the accuracy of forecast average accuracy is also improvedcompared to the commonly used Weighted Markov Chain.(3) Gray-fuzzy weight Markov chain model sensitive factor analysis, thecompletely fuzzy partition or the fuzzy division (WMCP) get high approximationcoefficient and Fuzzy Grouping. Another model is superimposed a weighted predictionresult of step4or step5is better than the1-3Step weighted effect.(4) Gives a fuzzy-order Markov chain is preferably a single interval proportionparameters and the model prediction method MATLAB calculation, gray-fuzzy weight the MATLAB calculation process of the Markov chain model. Even the field ofgeotechnical monitoring and analysis provides a good practical value for futurelandslide displacement deformation prediction. |
